Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
COLUMNS
BY
Zeynep Firat Alemdar
Chairperson
Committee Members*:
Committee:
_________________________
Chairperson
_________________________
_________________________
__________________________
__________________________
ii
ABSTRACT
iii
Two plastic hinge length expressions for reinforced concrete bridge columns
under static and dynamic loadings have been developed by studying the available test
results in the literature. Many of the previous tests were conducted using the static
loading and for small-scale components. A few of the tests focused on bridge columns
and dynamic loading. Expressions that have been developed to estimate the plastic hinge
lengths have either been based on the maximum drift at the top of the column, or the
spread of plasticity in the hinging regions. An expression to calculate the maximum drift
capacity of a bridge column in double curvature has been derived by considering the
deformations due to flexure as influenced by the definition of plastic hinge length (lp),
and the bond-slip effect of the longitudinal reinforcement at the connections. Drift
capacity of a bridge column, which corresponds to a 20% reduction in lateral load
capacity on the descending branch of the response backbone curve, has been estimated
using the new expression and compared with the results that were obtained from the
earlier plastic hinge length expressions. The measured drift of the bridge column from
the four-span large-scale bridge system test was also compared with the calculated
responses from the expressions. The proposed equations provide the best estimate of
plastic hinge length for reinforced concrete bridge columns.
iv
ACKNOWLEDGEMENTS
TABLE OF CONTENTS
GENERAL............................................................................................................. 1
1.2
1.3
1.4
1.3.1
1.3.2
1.3.3
1.3.4
1.3.5
1.3.6
1.3.7
1.3.8
1.3.9
Soesianawati, Park and Priestley (1986), Watson and Park (1994) ........ 21
1.3.10
1.3.11
1.3.12
1.3.13
1.3.14
1.3.15
1.3.16
TESTS) ........................................................................................................................... 49
1.4.1
1.4.2
1.5
SUMMARY......................................................................................................... 53
1.6
vi
DEFORMATIONS......................................................................................................... 58
2.1
INTRODUCTION ............................................................................................... 58
2.2
PROOF-OF-CONCEPT TEST............................................................................ 59
2.3
2.4
2.3.1
Description of Specimen.......................................................................... 67
2.3.2
Experimental Setup.................................................................................. 69
2.3.3
Earthquake Loading................................................................................. 74
2.3.4
Results...................................................................................................... 76
2.3.4.1
2.3.4.2
Displacements........................................................................................ 109
2.3.4.3
CONCLUSION.................................................................................................. 130
vii
iv
Figure 1.1 Curvature and deflection relationships for a reinforced concrete cantilever
(Paulay and Priestley 1992). ................................................................................................3
iv
Figure 1.2 Dimensions, steel content, and steel strain locations (Ernst 1957). ...................7
Figure 1.3 Exact curvature distributions for deflection calculations (Priestley and Park
1987) ..................................................................................................................................13
Figure 1.4 Effects of Various Parameters on Plastic Hinge Lengths (Sakai and Sheikh
1989) ..................................................................................................................................17
Figure 1.7 Relationship between Plastic Hinge Length and Shear Span-to-Depth Ratio
(Bae 2005) .........................................................................................................................35
Figure 1.8 Effect of Axial Load on Curvature and Compressive Strain Profiles (Bae
2005) ..................................................................................................................................36
Figure 1.9 Relationship between Plastic Hinge Length and Axial Load (Bae 2005)........37
Figure 1.11 Comparisons of Plastic Hinge Length (Eq. (1.20) versus Analysis) (Bae
2005) ..................................................................................................................................39
Figure 1.12 Idealization of curvature distribution in column: (a) column; (b) BMD; (c)
curvature diagram; and (d) equivalent curvature diagram (Restrepo et al. 2006) .............41
vi
Figure 1.14 Strain penetration coefficient (Restrepo et al. 2006)...................................44
Figure 1.15 Equivalent plastic hinge length as ratio of column diameter (Restrepo et al.
2006) ..................................................................................................................................44
Fig. 1.16 Simulated and observed force-drift ratio for columns with different design
parameters (Berry et al. 2008) ...........................................................................................49
Fig. 2.6 Displacement @ Grid Level A vs. Time (Rinaldi 0.95g) (1 in. = 254 mm) ........66
vii
Fig. 2.13 Close view of (a) bottom and (b) top grid systems ............................................71
Fig. 2.15 Metal weight pieces placed on each side of the tower .......................................72
Fig. 2.16 Location of the Bent 3 east column and the aluminum tower............................73
Fig. 2.17 Grid system and LVDT locations on column in the Bent 3 east column...........74
Fig. 2.18 Fix point on the wall at the back of the column .................................................77
Fig. 2.19 Lateral displacement of fixed point on the wall at Test 4D (1 in. = 254 mm) ...78
Fig. 2.20 Vertical displacement of fixed point on the wall at Test 4D (1 in. = 254 mm)..78
Fig. 2.21 Point 7 vertical displacement at the bottom grid system (1 in. = 254 mm)........82
Fig. 2.22 Five second interval to compare even- odd- and combination-line analyses (1
in. = 254 mm) ....................................................................................................................83
Fig. 2.23 Comparison of Point 7 vertical displacement for even lines with LVDT 3EBR7
data (1 in. = 254 mm) ........................................................................................................84
Fig. 2.24 Comparison of Point 7 vertical displacement for odd lines with LVDT 3EBR7
data (1 in. = 254 mm) ........................................................................................................85
vii
viii
Fig. 2.25 LVDT 3EBR7 vertical displacement history (1 in. = 254 mm) .........................85
Fig. 2.26 Rotation of horizontal line calculated from Points 7 and 8 at the bottom grid in
the transverse direction ......................................................................................................86
Fig. 2.27 Close up of lines used to define Point 3 (Fig. 2.17) ...........................................87
Fig. 2.29 Comparison of Point 7 vertical displacement with LVDT 3EBR7 data (1 in. =
254 mm).............................................................................................................................89
Fig. 2.30 Comparison of Point 9 vertical displacement with LVDT 3EBR8 data (1 in. =
254 mm).............................................................................................................................89
Fig. 2.33 Four corners surrounding general location of Point 12 and 13 ..........................92
viii
ix
Fig. 2.39 Point 7 vertical displacement vs. LVDT 3EBR7 data (1 in. = 254 mm)............97
Fig. 2.40 Point 9 vertical displacement vs. LVDT 3EBR8 data (1 in. = 254 mm)............97
Fig. 2.41 Rotation of horizontal line obtained using the constant Robert threshold .........98
Fig. 2.42 Point 7 vertical displacement compared with LVDT 3EBR7 data (1 in. = 254
mm)....................................................................................................................................98
Fig. 2.43 Point 9 vertical displacement compared with LVDT 3EBR8 data (1 in. = 254
mm)....................................................................................................................................99
Fig. 2.44 Rotation of horizontal line at h=7.7 in. (196 mm) from the bottom fixity.........99
Fig. 2.45 Comparison of Point 7 vertical displacement with LVDT 3EBR7 data (1 in. =
254 mm)...........................................................................................................................101
Fig. 2.46 Comparison of Point 9 vertical displacement with LVDT 3EBR8 data (1 in. =
254 mm)...........................................................................................................................101
Fig. 2.48 Comparison of Point 46 vertical displacement with LVDT 3ETR4 data
using
ix
Fig. 2.49 Comparison of Point 46 vertical displacement with LVDT 3ETR4 data using
the Edge Lines method (1 in. = 254 mm) ........................................................................104
Fig. 2.50 Rotation of vertical line calculated using Point 3 and 8...................................106
Fig. 2.53 Rotation of vertical line calculated using Point 38 and 52...............................107
Fig. 2.56 Comparison of Drift at Point 59 for Test 2 (1 in. = 254 mm) ..........................110
Fig. 2.57 Comparison of Drift at Point 59 for Test 4D (1 in. = 254 mm) .......................111
Fig. 2.58 Comparison of Drift at Point 59 for Test 6 (1 in. = 254 mm) ..........................111
Fig. 2.59 Maximum lateral movement ratios between photogrammetry and LVDT results
. ........................................................................................................................................112
Fig. 2.60 Overall deformed shape along the column height at maximum column drift
. ........................................................................................................................................114
Fig. 2.61 Top grid deformed shape with picture comparison at maximum column drift 115
xi
Fig. 2.63 The overall deformed shape along the column height at maximum column drift
. ........................................................................................................................................117
Fig. 2.64 Top grid deformed shape with picture comparison at maximum column drift
. ........................................................................................................................................118
Fig. 2.65 Bottom grid deformed shape with picture comparison at maximum column drift
. ........................................................................................................................................118
Fig. 2.66 The overall deformed shape along the column height at maximum column drift
. ........................................................................................................................................120
Fig. 2.69 Rotation of vertical line calculated using Point 3 and 8 for Test 2 ..................123
Fig. 2.70 Rotation of vertical line calculated using Point 8 and 13 for Test 2 ................123
Fig. 2.71 - Rotation of vertical line calculated using Point 38 and 45 for Test 2............124
Fig. 2.72 - Rotation of vertical line calculated using Point 45 and 52 for Test 2............124
Fig. 2.73 - Rotation of vertical line calculated using Point 3 and 8 for Test 4D .............125
Fig. 2.74 Rotation of vertical line calculated using Point 8 and 13 for Test 4D .............126
Fig. 2.75 Rotation of vertical line calculated using Point 38 and 45 for Test4D ............126
Fig. 2.76 Rotation of vertical line calculated using Point 45 and 52 for Test4D ............127
Fig. 2.77 Rotation of vertical line calculated using Point 3 and 8 for Test 6 ..................127
xi
xii
Fig. 2.78 Rotation of vertical line calculated using Point 8 and 13 for Test 6 ................128
Fig. 2.79 Rotation of vertical line calculated using Point 38 and 45 for Test 6 ..............128
Fig. 2.80 Rotation of vertical line calculated using Point 45 and 52 for Test 6 ..............129
Fig. 2.81 Maximum rotation ratios between photogrammetry compared LVDT results
. ........................................................................................................................................130
xii
xiii
LIST OF NOTATION
p : Plastic curvature
m : The maximum curvature
y : Curvature at first yield
u : The ultimate curvature
lp : The plastic hinge length
l : The height of a column
p : The plastic displacement
d : Distance from extreme compression fiber to centroid of tension reinforcement, in.
(mm)
w : Uniformly distributed load at a section of maximum moment kip/in., (kN/mm)
Vz : Shear adjacent to a concentrated load or reaction at a section of maximum moment
kips, (kN)
Mm : Maximum moment in a length of member kip-in., (kN.mm)
Me : Elastic-limit resisting moment kip-in., (kN.mm)
Mu : Ultimate resisting moment concurrent with Pu kip-in., (kN.mm)
Pu : Ultimate resisting axial load kips, (kN)
cue : Elastic component of cu, either calculated or assumed in the range 0.001 to 0.002,
cu : Maximum compressive strain in concrete at Mu and Pu, and
cuo : Basic maximum compressive strain in concrete
p : The plastic rotation,
k1 : The influence of the type of reinforcing steel,
k2 : The influence of column load (when present)
k3 : The influence of strength of concrete
xiii
xiv
xiv
INTRODUCTION
GENERAL
The determination of magnitude and location of inelastic deformations in
reinforced concrete bridge columns is a critical step for characterizing the performance
of the bridge system in earthquake events. Although it is possible to believe that some
ductility will be provided by beam hinges in bridge systems, it is generally the columns
of the bridges that must have inelastic rotational capacity. Bridge systems are designed to
keep inelastic behavior in the columns and away from the superstructure, which is
different than building systems.
If inelastic flexural deformations occur in a reinforced concrete structure due to
gravity and lateral loads, these deformations can generally be considered as concentrated
over a finite hinge length. The spread of plasticity, or hinge length is an important factor
in the analysis of deformation in reinforced concrete bridge structures and includes
elastic, plastic and softening stages of response. Little work has been completed to
determine hinge lengths in the plastic and softening phases of full-scale reinforced
concrete bridge systems. Previous work has relied on determining hinge lengths for
small-scale component tests. An evaluation of long-scale system behavior, as described
in this study, has the benefit of including the effects of moment redistribution and
progression of yielding throughout the entire structure.
1.2
nominal bending moment associated with yielding of the section. The equivalent plastic
hinge length can be calculated based on integration of the curvature distribution for
typical members. To simplify the calculations, an equivalent plastic hinge length lp can
be defined over which the plastic curvature, p, is assumed equal to m- y, where m is
the maximum curvature and y is the yield curvature, as shown in Fig. 1.1(a). The length
lp is determined so that the plastic displacement at the top of the cantilever column, p,
predicted from a displacement design method or from an experiment is the same as that
obtained from the actual curvature distribution as shown in Fig. 1.1(a). The lumped
plastic rotation, p, along the plastic hinge length is then computed as Eq. 1.1:
p = pl p = (m y ) l p
(1.1)
The plastic rotation derived using Eq. (1.1) can be used to determine the
displacement capacity of a section that experiences inelastic deformations. If the plastic
rotation is assumed to be concentrated at midheight of the plastic hinge, the plastic
displacement at the top of the cantilever column then becomes Eq. (1.2):
p = p (l 0.5l p ) = ( m y ) l p (l 0.5l p )
(1.2)
where l is the height of the column. The maximum nonlinear drift is then obtained from
the plastic displacement at the top of the cantilever. Therefore, a consistent prediction of
a plastic hinge length is necessary to examine the theoretical drift capacity of bridge
columns.
The hinge length indicates the theoretical length of damage concentration along
the column. Although the plastic hinge length should not be considered the required
column confinement region, it does indicate the minimum theoretical dimension. The
actual confinement region, which was shown as the extent of plasticity in Fig. 1.1 (a),
should be longer than the plastic hinge length.
y p
Extent of
plasticity
lp
(b) Deflections
Figure 1.1 Curvature and deflection relationships for a reinforced concrete cantilever
(Paulay and Priestley 1992).
1.3
researchers (Baker 1956, Baker and Amarakone 1964, Mattock 1964, 1967, Corley 1966,
Park, Priestley, and Gill 1982, Priestley and Park 1987, Paulay and Priestley 1992,
Sheikh and Khoury 1993, Mendis 2001). A detailed review of formulae that are available
to calculate the plastic hinge length for reinforced concrete columns are presented in this
section. Few studies, however, have been conducted to determine the plastic hinge length
in reinforced concrete bridge columns. These studies include Park, Priestley, and Gill
1982, Priestley and Park 1987, Tanaka and Park 1990, Watson and Park 1994, Kovacic
1995, Dodd et al. 2000, Hachem et al. 2003, Bae 2005, and Phan 2007, and are also
included in this section. In the dynamic tests, the equation proposed by Priestley et al.
(1992) was used to estimate the plastic hinge length before testing the bridge columns,
even though this equation was derived using the static test results of columns. Two
studies have been done to determine the plastic hinge length with dynamic testing of
3
reinforced concrete bridge column under dynamic base excitation (Dodd et al. 2000, and
Hachem et al. 2003) and these will be discussed later in this chapter.
1.3.1
The Institute of Civil Engineers committee published a report on the ultimate load
design of concrete structures (I.C.E. 1962), which includes the principles of ultimate load
theory and its application to design.
z
= k1k 2 k 3 ( ) 0.25
d
d
(1.3)
where k1, k2, and k3 represents the influence of the type of reinforcing steel, column load
(when present) and strength of concrete in lp respectively, z is the distance of critical
section to point of contraflexure, and d is the effective depth. The coefficients k1, k2, and
k3 were determined by examining several series of test results as described next.
From the analysis of test results that are described in this section (Ernst 1957,
McCollister et al. 1954, Poologasoundranayagam 1960, and Chan 1955), conservative
limiting values for k1, k2, and k3 are given as
0.7
k1 =
0.9
k2 = 1+0.5
0.6
k3 =
0.9
mild steel
(1.4)
(1.5)
cu = 6, 000 lb / in 2 (42 MPa )
(1.6)
where P is the column load acting in conjunction with bending moment, Pu is the load
capacity as an axially loaded column, and cu is the cube strength of concrete.
Ernst (1957) tested 30 simply-supported beams having 6 in. (150 mm) by 12 in.
(305 mm) cross-section, 10 ft. (3050 mm) in length, and a span of 9 ft. (2750 mm) under
central point loading. The main parameters were widths of column stubs and tension
steel as shown in Fig. 1.2. Column stubs changed in width from 6 to 36 in. (150 mm to
915 mm) at mid-span. The range of tension steel reinforcement ratios was from 1% to
5%. The nominal 28-day cube strength of the concrete was 3000 psi (22 MPa) or 4000
psi (28 MPa). Steel yield strength was approximately 45 ksi (310 MPa).
McCollister et al. (1954) designed 31 beams with 6 in. (150 mm) by 12 in. (305
mm) cross-section and 9 ft. (2750 mm) span. The main variables comprised the concrete
strength from 1905 psi (13 MPa) to 6407 psi (44 MPa), tension and compression steel
content (from 0.17% to 5.10% and from 0% to 4.08%, respectively), and the effect of
column stubs with dimensions of 6 in. (150 mm) by 6 in. (150 mm) cross-section and 12
in. (305 mm) height.
Poologasoundranayagam (1960) tested 38 simply supported beams having 4 in.
(100 mm) by 9 in. (230 mm) cross section and spans of 4-6 ft. (1220-1830 mm) under
central point loading. The principal factors were strength of concrete from 2,385 psi (16
MPa) to 6,330 psi (44 MPa), percentage of tension reinforcement (0.62% to 5.1%), and
quality of steel (mild or cold-worked and post-tensioned high tensile wire).
Chan (1955) conducted three series of tests. The first series consisted of nine short
prisms having 6 in. (150 mm) square sections and 11 in. (290 mm) long. They were
reinforced with 4 5/8 in. diameter (16 mm) longitudinal bars and used ties. These prisms
were pin-ended and loaded under compressive load with an eccentricity of in. (13
mm). Seven cylinders of 6 in. (150 mm) diameter and 12 in. (305 mm) length were
tested in the second series. They had the same longitudinal reinforcement with the first
series, but they were loaded in compression with an eccentricity of 1/4 in. (7 mm). The
main variable was lateral binding, where spiral reinforcement of in. (7 mm), 3/16 in.
(5 mm), and 1/8 in. (3 mm) diameter bars and pitches from 1 in. (25 mm) to 4 in. (100
mm) were used to confine the specimens. The last series included seven struts with 6 in.
(105 mm) by 3-5/8 in. (90 mm) cross section and 52 in. (1320) long. They were
symmetrically reinforced with 4 1/2 in. diameter (13 mm) bars, laterally bound with
rectangular welded closed links spaced at 3 in. (75 mm). They were tested under axial
compression through pins at the ends, and a central lateral point load. It was intended to
simulate a plastic hinge formation within the critical region of a column under bending
moment and high axial load. The nominal 28-day cube strength of the concrete was 3000
psi (22 MPa) for the first and second series, and 4000 psi (28 MPa) for the last series.
Average yield strength of the steel was around 40 ksi (275 MPa).
Chan (1962) evaluated the methods and parameters that were recommended in
the ICC report (1962) by studying thirteen column tests, comprising six by
Poologasoundranayagam (1960) and seven by Chan (1955), covering a range of P/Pu
from 0.06 to 0.78, cube strengths from 2,380 (16 MPa) to 5,160 psi (36 MPa), and
symmetrically reinforced steel ratios from 1.23% to 1.92%. Tests were analyzed and
compared in order to calculate values of EIe and p (plastic rotation) as described below.
Figure 1.2 Dimensions, steel content, and steel strain locations (Ernst 1957).
The testing procedure was similar to the study of 54 beam tests done by Chan
(1954); in addition, values of k2 were included to consider the effect of the column
loading. The histograms of the column test results were expressed as the ratio of
experimental to calculated values of EIe/ EIe and p/ p. They had a similar distribution
with the beam test results. Chan observed that the parameters used in Eq. (1.3) were safe
and statistically acceptable, however, the number of test results was small and more tests
were desired. The author also reported that in a broad range of structural members lp
varies from about 0.4 to 2.4d.
1.3.2
The ACI-ASCE Committee 428 on Limit Design (1968) recommended upper and
lower plastic hinge limits rather than a single equation. The length along a member from
the section of maximum moment, lp, should be bigger than the lesser of the two values
given in Eq. (1.7) and the value given in Eq. (1.8):
d
R + 0.03 zRm
4
and
Rd
(1.7)
(1.8)
in which;
R =
z=
0.004 cue
,
cuo cue
(1.9)
4M m
(1.10)
4V z + wM m Rm
Rm =
Mm Me
Mu Me
(1.11)
where;
1.3.3
and 10.8 ft. (3300 mm) in height were tested by Park et al. (1982). The longitudinal
9
10
Axial Load
Longitudinal
Reinf.
fc,
ksi
(Mpa)
(1)
3.35
(23)
P,
fy,
kips
ksi
l,
P/fcAg (Mpa)
%
(kN)
(3)
Unit
(5)
(2)
(4)
408.03
55.1
1
0.26
1.79
(1815)
(380)
602.49
55.1
2
6 (41)
0.214
1.79
(2680)
(380)
3.1
611.25
55.1
3
0.42
1.79
(21)
(2719)
(380)
3.41 958.81
55.1
4
0.6
1.79
(24)
(4265)
(380)
1: Compressive cylinder strength of concrete
Transverse Reinf.
db,
in.
(mm)
(6)
0.39
(10)
0.47
(12)
0.39
(10)
0.47
(12)
fyh,
ksi
(Mpa)
(7)
43.07
(297)
45.83
(316)
43.07
(297)
42.64
(294)
s,
%
(8)
Ash/Ash,ACI
(9)
1.5
0.66
2.3
0.63
0.89
3.5
1.47
Meas.
Lp,
in.
(mm)
(10)
9.53
(242)
7.44
(189)
8.62
(219)
10.75
(273)
Lp/h
(11)
0.44
0.34
0.4
0.5
1.3.4
Mander (1983)
Experimental studies conducted by previous investigators at the University of
Canterbury (Gill et al. 1979, Potangaroa et al. 1979, Ghee et al. 1981, Davey et al. 1975,
Munro et al. 1976 and Heng et al. 1978) have supported the theory that the equivalent
plastic hinge length, lp, may vary from 0.35 to 0.65 of the overall member depth for solid
reinforced concrete columns. Based on a comparison of the available results for
octagonal specimens (RRU 1983), it was found that the equivalent plastic hinge length is
11
independent of the axial load level and aspect ratio. A value of lp = 0.5D was
recommended.
After examining the experimental results studied at the University of Canterbury,
Mander concluded that contributions to plastic deformation were primarily from two
sources: (i) the spread of plasticity along the member length due to the moment gradient
and (ii) yield penetration of the longitudinal reinforcement beyond the limits of the
plastic hinge. The equivalent length of yield penetration, Lpy, could be written in terms of
the longitudinal bar diameter from the force-deflection analyses:
L py = 6.35 d b
(in.)
(1.12a)
L py = 32 d b
(mm)
(1.12b)
(1.13)
When the predicted and observed results are compared, Eq. (1.13) generally
provides a conservative prediction of the equivalent plastic hinge length. Mander also
noted that Eq. (1.13) must not be used for estimating the length requiring detailed
confinement because plastic curvature would spread over approximately three equivalent
plastic hinge lengths.
12
1.3.5
distribution along the column, an alternate approach was developed by Priestley and Park
(1987) considering the moment-curvature relationships for different sections along the
height of the column. The curvature distribution along the column can be calculated
using Eq. (1.14) for any given base moment as shown in Fig. 1.3(a). The predicted
displacement at the top of the column is then obtained by integrating the curvature
profile.
L
= ( x ) xdx
(1.14)
(a)
(b)
Figure 1.3 Exact curvature distributions for deflection calculations (Priestley and Park
1987)
13
14
were tested under axial load only. The range of longitudinal reinforcement ratios was
between 1% and 4%, and the lateral reinforcement ratio was from 0.5% to 1.5% with
spiral or circular hoops. The axial load values ranged from 0.2Po to 0.7Po, where
Po=Pe/fcAg (Pe is design compressive load of the column due to gravity and seismic
loading, fc is compressive cylinder strength of concrete, and Ag is gross area of section).
In the second stage, the sections included square, diagonal, octagonal, and hollow square
shapes were tested under continued axial load and cyclic reversals of bending moment.
The test columns were instrumented extensively along the potential plastic hinge regions.
Priestley and Park (1987) proposed a general plastic hinge length formula (Eq.
1.15) based on the new test results.
l p = C1 L + C 2 d b + C 3 D
(1.15)
where L is the distance from the point of contraflexure of the column to the section of
maximum moment, db is the longitudinal bar diameter, D is section depth (or diameter
for circular sections) and C1, C2, and C3 are constants determined from curvature
distributions along the column length of the specimen.
Curvature distributions along the length were obtained for all units during the column
tests to predict the values of the constants. Best fit values of C1 = 0.08, C2 = 6, and C3 =
0 were found based on the analysis of the test results. Therefore, Eq. 1.15 becomes
l p = 0.08 L + 6d b
(ksi)
(1.16a)
l p = 0.08 L + 0.88d b
(MPa)
(1.16b)
Priestley and Park (1987) obtained good agreement between the experimentally
derived values for lp and values calculated using Eq. (1.16). The tests that were evaluated
included studies outside this program (Gill et al. 1982, Potangaroa et al. 1981, Davey et
al. 1975, Munro et al. 1976, Ng et al. 1978, Ghee et al. 1985 and Mander et al. 1984).
15
The average hinge length that was calculated for all tests was approximately equal to lp=
0.5D as shown in Table 1.2. The experimental data did not show any relationship
between plastic hinge length and axial load ratio, longitudinal reinforcement ratio, or
yield stress of longitudinal reinforcement.
Researchers
Davey et al.
Munro et al.
Heng et al.
Gill et al.
Potangaroa et al.
Ghee et al.
Mander et al.
Column
aspect ratio
4
4
5.5
4
2.18
2.18
2.18
2.18
2
2
2
2
4
4
4
4
4.27
4.27
4.27
4.27
16
1.3.6
hinge length increased as the aspect ratio (L/h or L/D, which is equivalent to the shear
span-to-depth ratio for cantilever columns) increased. Bilinear curves were developed to
give the relationship between the plastic hinge length and the aspect ratio as shown in
Figure 1.4. They concluded that the amount of transverse reinforcement, axial load level,
and aspect ratio had an influence on the plastic hinge length. The plastic hinge length
generally increased with increasing values of each parameter.
Figure 1.4 Effects of Various Parameters on Plastic Hinge Lengths (Sakai and Sheikh
1989)
1.3.7
four column specimens (Units 1 to 4) had a total height of 5.9 ft. (1800 mm) and 16 in.
17
(400 mm) square sections. In the second series, four column specimens (Units 5 to 8) had
a total height of 5.4 ft. (1650 mm) and 22 in. (550 mm) by 22 in. (550 mm) square
sections. The shear span-to-depth ratios were 4 and 3 for the first and second series,
respectively. The level of applied axial load (P/fcAg), the shear span-to-depth ratio of
each column (L/h), the configuration of transverse reinforcement, and anchorage details
of that reinforcement were the main variables. Table 1.3 illustrates the mechanical
properties of the materials and other details of the column specimens.
The plastic hinge region of the column specimens was designed according to the
NZS 3101:1982 code for both confinement and shear. After testing the two series of
column units, the equivalent plastic hinge lengths were found to be between 0.40 and
0.75 of the overall depth of the column section. Tanaka and Park (1990) observed that
when the axial load level increased, the equivalent plastic hinge length increased.
18
Transverse Steel
fyh,
ksi
(MPa)
(6)
48.29
(333)
48.29
(333)
48.29
(333)
48.29
(333)
47.13
(325)
47.13
(325)
47.13
(325)
47.13
(325)
s
(%)
(7)
Ash/Ash,ACI
(8)
bxh,
in. (mm)
(9)
L/h
(10)
2.55
1.06
15.75x15.75
4.00
2.55
1.06
15.75x15.75
4.00
2.55
1.06
15.75x15.75
4.00
2.55
1.06
15.75x15.75
4.00
1.70
0.82
21.65x21.65
3.00
1.70
0.82
21.65x21.65
3.00
2.08
1.00
21.65x21.65
3.00
2.08
1.00
21.65x21.65
3.00
Lp,
in.
(mm)
(11)
6.77
(172)
8.7
(221)
10.6
(269)
11.06
(281)
8.35
(212)
13.66
(347)
14.45
(367)
18.66
(474)
1.3.8
plastic hinge length based on integration of the curvature distribution for typical
members should be dependent on l, where l is the column height. Plastic hinge length
values that are needed to calculate a measured lateral displacement, however, were not
19
consistent with experimentally measured lengths of lp. As Fig. 1.5 shows, the theoretical
curvature distribution ends abruptly at the base of the cantilever (Fig. 1.5b), whereas the
actual steel tensile strains should continue for some depth into the footing due to finite
bond stress. The elongation of longitudinal bars beyond the theoretical base causes
additional rotation and deflection (tensile strain penetration as shown in Fig. 1.5c). The
following formula was proposed by revising Eq. (1.16) to consider the effect of flexural
reinforcement with different strengths on the length of a plastic hinge formed at the
bottom of a cantilever column:
l p = 0.08l + 0.15d b l f ye 0.3d b l f ye
(ksi)
(1.17a)
(MPa)
(1.17b)
where l is the height of the cantilever column, fye is the yield stress of longitudinal
reinforcement, and dbl is the diameter of the longitudinal reinforcement.
Paulay and Priestley (1992) recommended that Eq. (1.17) results in values of lp
close to 0.5d, where d is the section depth, for typical beams and columns. It was
observed that the equivalent plastic hinge length and the region of plasticity where
special reinforcing detailing is required must be defined separately to ensure dependable
inelastic rotation capacity. This difference is shown in Fig. 1.5b by indicating the spread
of plasticity over a region outside the equivalent plastic hinge length.
20
(b) Curvature at
maximum response
Figure 1.5 Theoretical curvature relationships for a prismatic reinforced concrete
cantilever column (Paulay and Priestley 1992)
1.3.9
concrete columns under low axial loads. The column specimens were designed with
smaller quantities of confining reinforcement than those recommended by the
NZS3101:1982 code (New Zealand Standards). Watson and Park (1994) furthered the
experimental research of Soesianawati et al. (1986) by testing five more square columns
and two octagonal columns under moderate to high axial compression load levels. Table
1.4 gives the details of the square column specimens, which have 16 in. (400 mm) square
cross sections and a height of 64 in. (1600 mm). The shear span-to-depth ratio was 4 for
the test specimens.
Units 1 to 4 were subjected to low axial load (P = 0. 1fcAg to 0. 3fcAg). Units 1
and 2 contained 43% and 46% of the New Zealand code recommended quantity of
transverse reinforcement. These specimens reached displacement ductility factors of at
21
22
lc
P
= 1 + 0.4 '
h
f c Ag
(ksi)
(1.18a)
lc
P
= 1 + 2.8 '
h
f c Ag
(MPa)
(1.18b)
where
lc = length of confined region of column, in. (mm)
h = lateral dimension of rectangular column section, in. (mm)
23
Longitudinal
Reinforcement
fc ,
P,
fy,
ksi
kips
ksi
(MPa) (kN) P/fcAg (MPa)
(1)
(2)
(4)
Unit
(3)
6.74
167
64.67
1
0.1
(46)
(743)
(446)
6.38
475
64.67
2
0.3
(44)
(2113)
(446)
6.38
475
64.67
3
0.3
(44)
(2113)
(446)
5.8
432
64.67
4
0.3
(40)
(1922)
(446)
5.95
737
68.73
5
0.5
(41)
(3278)
(474)
5.8
719
68.73
6
0.5
(40)
(3198)
(474)
6.09
1058
68.73
7
0.7
(42)
(4706)
(474)
5.65
982
68.73
8
0.7
(39)
(4368)
(474)
5.8
1007
68.73
9
0.7
(40)
(4479)
(474)
1: Compressive cylinder strength of concrete
l
(%)
(5)
1.51
1.51
1.51
1.51
1.51
1.51
1.51
1.51
1.51
Transverse Reinforcement
d b,
in.
(mm)
(6)
0.28
(7)
0.32
(8)
0.28
(7)
0.24
(6)
0.32
(8)
0.24
(6)
0.47
(12)
0.32
(8)
0.47
(12)
fyh,
ksi
(MPa)
(7)
82
(565)
81
(558)
82
(565)
57
(393)
84
(579)
87
(600)
69
(476)
84
(579)
69
(476)
s
(%)
(8)
Ash/Ash,ACI
(9)
lp/h
(Eq.
1.13)
(10)
0.84
0.36
0.56
1.2
0.55
0.56
0.79
0.36
0.56
0.56
0.2
0.56
1.15
0.58
0.56
0.55
0.29
0.56
2.16
0.9
0.56
1.21
0.64
0.56
3.99
1.75
0.56
1.3.10 Sheikh and Khoury (1993), Sheikh, Shah and Khoury (1994)
Sheikh and Khoury (1993) and Sheikh et al. (1994) completed experimental
research on six large-scale normal-strength concrete and four high-strength concrete
column specimens. The concrete columns were 72.5 in. (1842 mm) high and had 12-in.
(305 mm) square cross sections that result in a shear span-to-depth ratio of 6. The
24
concrete strength, level of axial load, and the percentage of transverse reinforcement
were the main test variables. Table 1.5 shows the details of the specimens tested and the
applied axial load.
The primary goal of the research was to assess the confinement provisions of the
ACI 318 (1989) code. According to this version of the code, the total cross sectional area
of rectangular hoop reinforcement for confinement (Ash) should not be less than that
given by the following Eq. (1.19a and b):
Ag
f c'
Ash = 0.3sbc
1
A
ch
f yt
Ash = 0.09 sbc
f c'
f yt
(1.19a)
(1.19b)
where
Ag = gross area of column section, in.2 (mm2)
Ach = area of core concrete measured out-to-out of transverse reinforcement, in.2
(mm2)
fc = compressive strength of concrete, ksi (MPa)
fyt = yield strength of transverse reinforcement, ksi (MPa)
s = spacing of transverse reinforcement, in. (mm)
bc = cross sectional dimension of column core, measured center-to-center of
transverse reinforcement, in. (mm).
In the ACI 318 code, the length of the column requiring confinement is specified
as the greatest of the overall depth (h) of a column at the joint face (where h is the larger
sectional dimension for a rectangular column or the diameter of a circular column), onesixth of the clear height of a column, or 18 in. (457 mm). The spacing of transverse
25
reinforcement is required to be less than h/4 or 6db, where h is the minimum member
dimension and db is the diameter of longitudinal reinforcement.
The researchers concluded that a column designed according to the ACI (1989)
code requirements has adequate performance in terms of curvature and displacement
ductility, but only for certain situations. Depending on the reinforcement detailing and
axial load level, the code provisions may give unnecessarily conservative design. It was
also observed that the measured plastic hinge lengths were an average value of 1.0h in
the column tests as shown in Table 1.5, where h is the column depth. Most of the column
tests were, however, conducted under high axial loads. It also appeared that steel
configuration, axial load level, amount of confining steel, and concrete strength did not
have an influence on the plastic hinge length.
26
No.
of
bars
(2)
FS-9
4.7
(32)
8
8
ES-13
4.72
(33)
AS-3
4.81
(33)
AS-17
4.54
(31)
4.75
(33)
Spec.
AS_18
4.68
(32)
l
(%)
(3)
fyl,
ksi
(MPa)
(4)
s
(%)
(5)
fyh,
ksi
(MPa)
(6)
2.44
73.6
(507)
1.68
2.44
73.6
(507)
P/fcAg
(8)
73.6
(507)
1.46
0.76
13.1
(333)
1.10
1.69
67.3
(464)
1.34
0.76
10.2
(259)
0.85
2.44
73.6
(507)
1.68
73.6
(507)
1.43
0.60
11.5
(292)
0.96
2.44
73.6
(507)
1.68
73.6
(507)
1.52
0.77
12.6
(320)
1.05
73.6
(507)
3.06
67.3
(464)
2.41
0.77
11.9
(302)
0.99
73.6
(507)
1.30
73.6
(507)
67
(462)
1.12
0.47
13.9
(353)
1.16
2.44
73.6
(507)
1.68
73.6
(507)
0.88
0.62
12.7
(323)
1.05
2.44
73.6
(507)
3.06
67.3
(464)
1.44
0.64
10.7
(272)
0.89
2.44
73.6
(507)
4.30
67.3
(464)
2.10
0.64
13
(330)
1.08
2.44
73.6
(507)
1.68
73.6
(507)
0.80
0.65
2.44
8
AS-19
2.44
AS3H
7.86
(54)
AS18H
7.93
(55)
AS20H
7.78
(54)
A-17H
8.57
(59)
Lp,
in.
(mm)
(9)
Ash/
Ash,ACI
(7)
Lp/h
(10)
27
Concrete
Span,
Width,
Depth,
Axial
strength
in.
in.
in.
force,
fc, ksi
Measured
ACI 1ACI 2(mm)
(mm)
(mm)
kips (kN)
Label
(MPa)
Lp/d
Lower/d
Upper/d
50
3.15
5.91
14.61
4.82
0.254
(1270)
(80)
(150)
(65)
(33)
D1
0.37
0.91
50
3.15
5.91
14.61
4.74
0.467
D2
(1270)
(80)
(150)
(65)
(33)
0.37
0.91
50
3.15
5.91
14.61
5.61
0.633
D3
(1270)
(80)
(150)
(65)
(39)
0.37
0.91
50
3.15
5.91
29.22
8.43
0.299
D4
(1270)
(80)
(150)
(130)
(58)
0.37
0.91
50
3.15
5.91
14.61
9.33
0.467
D5
(1270)
(80)
(150)
(65)
(64)
0.37
0.91
50
3.15
5.91
29.22
9.4
0.699
D6
(1270)
(80)
(150)
(130)
(65)
0.37
0.91
50
3.15
5.91
7.19
9.11
0.547
(1270)
(80)
(150)
(32)
(63)
D7
0.37
0.91
50
3.15
5.91
14.61
9.46
0.467
D8
(1270)
(80)
(150)
(65)
(65)
0.37
0.91
1: The ratio of plastic hinge length calculated using ACI lower limit (1968) to the depth of the column
2: The ratio of plastic hinge length calculated using ACI upper limit (1968) to the depth of the column
28
concrete columns on plastic hinge length. The concrete strength for standard cylinders
ranged between 10,000 and 16,000 psi (72 MPa and 112 MPa). The cross-sections of the
columns were 12 in. (305 mm) square, and 12 in. (305 mm) by 10 in. (250 mm)
rectangular dimensions with 72.5 in. (1,841 mm) in height. The shear span-to-depth
ratios were 6, 7.4 and 5.3.
The plastic hinge lengths of the specimens tested were calculated using the Eq.
(1.2) for all the load cycles in which the displacement ductility factor is greater than 4
and then averaged to find the equivalent plastic hinge length for the columns. The
experimental plastic hinge lengths were close to the depth of column sections (h) as
given in column (9) of Table 1.7. It was suggested that a simpler expression such as Lp =
x*h, where x can have a value between 0.9 and 1, is more appropriate to obtain the
plastic hinge length for the columns.
In the two studies, as the axial load increased, the deformability of the reinforced
concrete columns reduced and strength and stiffness degradation with every load cycle
accelerated. Thus, a larger amount of lateral reinforcement was needed to balance this
effect. Bayrak and Sheikh (1997) concluded that the axial load level should be
considered in the design of confining reinforcement. Based on the test results, the
displacement ductility factors decreased with increasing shear span-to-depth ratios (L/h).
It was observed that section geometry and shear span-to-depth ratio influenced the
member-level ductility parameters (which are the displacement ductility factor, and work
damage indicator that was represented by the work done on the column by lateral load),
whereas section-level ductility parameters (which are the curvature ductility factor, and
energy damage indicator that was defined by energy dissipated in the plastic hinge
region) were not affected by these factors.
29
Unit
ES-1HT
AS-2HT
AS-3HT
AS-4HT
AS-5HT
AS-6HT
AS-7HT
ES-8HT
RS-9HT
RS10HT
RS11HT
RS12HT
RS13HT
RS14HT
RS15HT
RS16HT
RS17HT
RS18HT
RS19HT
RS20HT
WRS21HT
WRS22HT
WRS23HT
WRS24HT
fc, ksi
(MPa)
(1)
10.45
(72)
10.4
(72)
10.41
(72)
10.43
(72)
14.76
(102)
14.78
(102)
14.79
(102)
14.82
(102)
10.32
(71)
10.31
(71)
10.27
(71)
10.28
(71)
16.25
(112)
16.25
(112)
8.15
(56)
8.15
(56)
10.74
(74)
10.74
(74)
10.76
(74)
10.76
(74)
13.24
(91)
13.24
(91)
10.47
(72)
10.47
(72)
P/fcAg
(2)
0.5
0.36
0.5
0.5
0.45
0.46
0.45
0.47
0.34
0.5
0.51
0.34
0.35
0.46
0.36
0.37
0.34
0.5
0.53
0.34
0.47
0.31
0.33
0.5
Longitudinal
Steel
fy, ksi
l
(MPa) (%)
(3)
(4)
65.83
(454) 2.58
65.83
(454) 2.58
65.83
(454) 2.58
65.83
(454) 2.58
65.83
(454) 2.58
65.83
(454) 2.58
65.83
(454) 2.58
65.83
(454) 2.58
65.83
(454) 2.74
65.83
(454) 2.74
65.83
(454) 2.74
65.83
(454) 2.74
65.83
(454) 2.74
65.83
(454) 2.74
65.83
(454) 2.74
65.83
(454) 2.74
75.55
(521) 2.74
75.55
(521) 2.74
75.55
(521) 2.74
75.55
(521) 2.74
75.55
(521) 2.74
75.55
(521) 2.74
75.55
(521) 2.74
75.55
(521) 2.74
Transverse Steel
Spec.,in. fyh, ksi
s
(mm)
(MPa) (%)
(6)
(5)
(7)
3.74
67.14
(95)
(463) 3.15
3.54
78.59
(90)
(542) 2.84
3.54
78.59
(90)
(542) 2.84
3.94
67.14
(100)
(463) 5.12
3.54
78.59
(90)
(542) 4.83
2.99
67.14
(76)
(463) 6.72
3.7
78.59
(94)
(542) 2.72
2.76
67.14
(70)
(463) 4.29
3.15
78.59
(80)
(542) 3.44
3.15
78.59
(80)
(542) 3.44
3.15
78.59
(80)
(542) 5.43
5.91
78.59
(150)
(542) 1.83
2.76
67.43
(70)
(465) 3.92
2.76
67.43
(70)
(465) 3.92
3.94
67.43
(100)
(465) 2.75
5.91
67.43
(150)
(465) 1.83
2.95
197.2
(75)
(1360) 1.83
2.95
197.2
(75)
(1360) 1.83
2.95
203.29
(75)
(1402) 3.54
5.51
203.29
(140)
(1402) 1.9
2.76
67.43
(70)
(465) 3.92
2.76
67.43
(70)
(465) 3.92
3.15
78.59
(80)
(542) 3.44
3.15
78.59
(80)
(542) 3.44
30
Section Exp.
Depth,
Lp,
in.
in.
Ash/
(mm)
(mm)
Ash,ACI
(9)
(10)
(8)
12.01 13.82
1.13
(305)
(351)
12.01 11.73
1.19
(305)
(298)
12.01 10.87
1.19
(305)
(276)
12.01 10.71
(305)
(272)
1.83
12.01 10.31
1.08
(305)
(262)
12.01 12.64
1.62
(305)
(321)
12.01 10.55
0.8
(305)
(268)
12.01 15.16
1.08
(305)
(385)
13.78 14.09
(350)
(358)
1.72
13.78 17.48
1.72
(350)
(444)
13.78 15.59
2.29
(350)
(396)
13.78 16.42
0.92
(350)
(417)
13.78 11.65
1.09
(350)
(296)
13.78 13.82
(350)
(351)
1.09
13.78 10.71
1.49
(350)
(272)
13.78 14.84
1
(350)
(377)
13.78 11.65
1.39
(350)
(296)
13.78 12.56
1.39
(350)
(319)
13.78 13.54
2.67
(350)
(344)
13.78 13.74
(350)
(349)
1.43
9.84
11.02
1.31
(250)
(280)
9.84
10.98
1.31
(250)
(279)
9.84
10.08
1.7
(250)
(256)
9.84
9.72
1.7
(250)
(247)
Average
Standard Deviation
Lp/h
(11)
1.15
0.98
0.91
0.89
0.86
1.05
0.88
1.26
1.02
1.27
1.13
1.19
0.85
1
0.78
1.08
0.85
0.91
0.98
1
1.12
1.11
1.03
0.99
1.01
0.13
Bar
Size,
(SI)
(3)
No.9
(29)
No.71
(22)
l
(%)
(4)
2.08
1.25
No.5
(16)
fyl,
ksi
(MPa)
(5)
84
(579)
73
(503)
s
(%)
(6)
fyh,
ksi
(MPa)
(7)
1.28
2.04
72
(496)
1.25
No.72
(22)
No.72
(22)
Transverse Steel
1.25
1.25
64
(441)
62
(427)
58
(400)
P/Po
(9)
1.04
0.5
1.09
0.5
1.12
0.5
0.44
0.2
0.74
0.2
72
(496)
1.76
58
(400)
Ash/
Ash,ACI
(8)
0.72
1.3
66
(455)
63
(434)
32
The behaviors that were observed from specimens S24-2UT and S17-3UT were
compared in detail to examine the effect of shear span-to-depth ratio (L/h) on the
response of concrete columns. Specimens S24-2UT and S17-3UT had 24 in. (610 mm)
and 17.25 in. (440 mm) square cross-sections, respectively. These two specimens were
designed with similar concrete strength, amount and detail of longitudinal and transverse
reinforcement. The only difference between them was the shear span-to-depth ratio for
specimens S24-2UT and S17-3UT due to different section sizes. The shear span-to-depth
ratios of these specimens were calculated as 5 and 7, respectively. Similar sectional
performance was obtained from tests of S24-2UT and S17-3UT. The curvature ductility
factors, which are the ratio of ultimate curvature to the curvature at first yield, were
33
calculated from the backbone curves of specimens S24-2UT and S17-3UT as 8.8 and
9.1, respectively. The member-level deformation capacities of the specimens, however,
were quite different. It was observed that as the shear span-to-depth ratio increased from
5 to 7, displacement ductility was significantly decreased from 5.9 to 2.7. The drift
capacity then dropped from 2.6% to 1.6%. From all the observations above, it can be
summarized that the shear span-to-depth ratio had an important effect on the overall
member performance of the test columns.
Bae (2005) introduced an analytical approach to estimate the plastic hinge length
of a concrete column. In this approach, first the intensity of concrete compressive strains
is measured at critical sections along the length of a column, and then using the
compressive strain profile, the yielding region length of longitudinal bars can be
estimated. Finally, the plastic hinge length is calculated by subtracting the stub
confinement effect (approximately 0.25h where h is the column depth) from the overall
length where the compressive bar strains are greater than the yield strain.
A series of theoretical analyses were carried out to study the influence of the
shear span-to-depth ratio on the plastic hinge length. A 24 in. (610 mm) square column
having different shear span-to-depth ratios (L/h) was designed. The longitudinal
reinforcement was constant as l = 0.01. The results of the analyses are shown in Fig.
1.7. The plastic hinge length increased with increasing L/h, and this effect was more
pronounced as the axial load increased.
34
Figure 1.7 Relationship between Plastic Hinge Length and Shear Span-to-Depth Ratio
(Bae 2005)
35
(displacement ductility, and energy absorption) for specimen S24-2UT gave higher
values than those of specimen S24-4UT, but less than those of specimen S24-5UT.
To determine the sensitivity of the analysis, a 24 in (610 mm) square column with
120 in. (3,050 mm) length was studied to investigate the effect of axial load on the
plastic hinge length. The percentage of longitudinal reinforcement was 1% for each
specimen. Concrete strength was 6,000 psi (42 MPa). The yield and ultimate strength of
reinforcing bars were 60 ksi (414 MPa) and 90 ksi (620 MPa), respectively. Along the
length of columns, the curvature and compressive concrete strain profiles were examined
for different axial load levels (from 0.2Po to 0.5Po).
Figure 1.8 shows the effect of axial load on curvature and compressive strain of
concrete. The curvature profiles are not affected from the various axial load levels
whereas the compressive strain profiles indicate the influence of axial load very well.
Figure 1.8 Effect of Axial Load on Curvature and Compressive Strain Profiles (Bae
2005)
To estimate the plastic hinge length, the region of reinforcing bars that yield in
compression was investigated along the column. Fig. 1.9 shows that the plastic hinge
36
length is nearly constant for low axial loads (P 0.2Po) where it is approximately equal
to h/4. The plastic hinge length increases as axial loads increase (P>0.2Po).
Figure 1.9 Relationship between Plastic Hinge Length and Axial Load (Bae 2005)
The estimated plastic hinge length is approximately equal to 0.8h at the axial load
level of 0.52Po, which is the maximum axial load allowed by ACI 318-05 (Pn,max =
0.8Po = 0.650.8Po=0.52Po for tie reinforcement). An additional length of 0.25h
should be added to the plastic hinge to take into account the stub confinement effect. The
required confined length was calculated as 1.05h by adding the largest estimated plastic
hinge length and a distance of 0.25h. The length of the region of confinement should be
the largest of the overall depth of a column, one-sixth of the clear height of a column, or
18 in. (457 mm) according to Chapter 21 of the ACI 318-05 code. The confinement
region length indicated by ACI 318-05 appears to be satisfactory for low axial loads.
To evaluate the effect of longitudinal reinforcement on the plastic hinge lengths,
a 24 in. (610 mm) square column with 120 in. (3,050 mm) length was designed. The
shear span-to-depth ratio of the column specimens was 5. The curvature and compressive
37
strain profiles along the length of the column were examined to obtain the plastic hinge
lengths. Results showed that as the longitudinal reinforcement ratio (l) increases, the
length of the plastic hinge increases. Figure 1.10 shows the results for different axial load
levels.
After all the sensitivity analyses were done, results showed that axial load, shear
span-to-depth ratio, and the amount of longitudinal reinforcement significantly
influenced the length of plastic hinges. Using linear relationships between the parameters
(P/Po, L/h, and As/Ag) and the plastic hinge, a plastic hinge length expression was
calibrated based on the analysis results. Equation 1.25 was proposed from the result of a
series of least squares analyses that were done on the UW/PEER column database
(http://maximus.ce.washington.edu/~peera1/).
P
= 0.3
h P0
lp
As
+ 3
Ag
38
(1.20)
The plastic hinge lengths obtained from the analysis of compressive strains along
the column height and those from using Eq. 1.20 were compared in Fig. 1.11. It was
observed that the plastic hinge lengths estimated by the analytical approach and Eq. 1.20
did not match very well, especially at high axial loads and large shear span-to-depth
ratios, because the plastic hinge lengths obtained from the analysis of compressive
strains are approximate results. It was important to mention that Bae (2005) proposed to
find a simple expression so that conservative estimations for the deformation capacity of
concrete columns can be obtained. It can be summarized that moderate plastic hinge
length values can be determined using the new formula for a wide range of axial loads,
shear span-to-depth ratios, and amount of longitudinal reinforcement.
Measured
Lp
Proposed
Lp
S24-1UT
S24-2UT
S17-3UT
S24-4UT
S24-5UT
0.66h
0.91h
0.49h
0.47h
0.69h
0.86h
0.25h
0.25h
(1.21)
40
(1.22)
The resulting deflection can be divided into elastic elf (Eq. 1.23a) and plastic pf
(Eq.1.23b) components that are consistent with the equivalent curvature distribution
drawn in Fig. 1.12(d). This calculation is demonstrated in Eq. (1.23):
M ' H2
y
M
y
elf =
(1.23a)
pf = pH 2 (1 )
(1.23b)
p =
M '
y
My
(1.24)
Combining Eq. (1.22) through (1.24), and finding the solution for
fle
= 1 1 2
(a)
M ' H2
(
y )
My
3
(1.25)
pH 2
(b)
(c)
(d)
Figure 1.12 Idealization of curvature distribution in column: (a) column; (b) BMD; (c)
curvature diagram; and (d) equivalent curvature diagram (Restrepo et al. 2006)
The strain penetration coefficient can be obtained from the reading of the first
and second displacement transducers that were instrumented immediately above the
41
column base. The fixed-end rotation shown in Fig. 1.12(a) can be calculated from the
first set of displacement transducers if the gauge length is small enough to minimize the
effect of local bond slip in the rotation. The second set of displacement transducers is
used to estimate the base curvature shown in Fig. 1.12(d). The plastic hinge length due
to strain penetration can be written as
f ye d bl =
(1.26)
Therefore, is
(1.27)
f ye d bl
Figure 1.13 and 1.14 illustrate and versus the curvature ductility demand
gained from both test units, respectively. The values of = 0.08 and = 0.15 (if fye is in
MPa, is equal to 0.022) used in Eq. (1.17) are shown in these figures.
It can be seen that the coefficient increases as the curvature ductility demand
increases because the spread of plasticity takes place gradually until a column reaches its
maximum ductility capacity. Figure 1.13 shows that the maximum spread of plasticity is
significantly greater than predicted by Paulay and Priestley (1992) for both test results.
Unlike the spread of plasticity coefficient , the strain penetration coefficient is
absolutely independent from the curvature ductility demand as shown in Fig. 1.14. It is
seen that the values of coefficient obtained for Units 1 and 2 are also greater than the
value proposed by Paulay and Priestley (1992). It is expected that equivalent plastic
hinge lengths will be greater than those obtained by Eq. (1.17) because both values of
coefficients are greater than the values recommended by Paulay and Priestley (1992).
Therefore, the theoretical lateral deformation capacity is greatly underestimated for
circular columns when using Eq. (1.17).
42
43
Figure 1.15 Equivalent plastic hinge length as ratio of column diameter (Restrepo et al.
2006)
44
1.3.15 Phan V., Saiidi M.S., Anderson J., and Ghasemi H. (2007)
Phan et al. (2007) designed two one-third scale reinforced concrete bridge
columns (labeled NF-1 and NF-2) that were tested on a shake table at the University of
Nevada, Reno Large Scale Structures Laboratory to investigate near-fault ground motion
effects on bridge columns. NF1 and NF2 had the same cross-sectional properties and
column height. The only difference was that NF1 was designed according to 2004
Caltrans Seismic Design Criteria (SDC) version 1.3, whereas NF2 was designed based
on the AASHTO 2002 Standard Specifications for Highway Bridges (AASHTO 2002).
NF1 column specimen was nearly identical to 9F1 that was tested in the same laboratory
during a previous study (Laplace 2005). The only difference between NF-1 and 9F1 was
that NF-1 was subjected to near-fault impulsive ground motion whereas 9F1 was tested
under an earthquake record that did not include forward directivity effects (El Centro
1940).
Paulay and Priestleys equation (Eq. 1.17) was used to estimate the theoretical
plastic hinge length (lp) for each specimen. The theoretical values are generally
conservative for conventional reinforced concrete columns. For all three specimens, the
theoretical lp was calculated as 11.26 in. (286 mm). Table 1.10 lists the measured lp for
all the test specimens to examine whether current methods for calculating lp may
underestimate the actual value in structures subjected to near-fault ground motions
(Hamilton et al. 2001). The measured lp were 51-95% longer than those determined using
Eq. (1.17). The lp for specimen NF-2 was very similar to that of specimen 9F1, and lp for
NF-1 was also comparable. Phan et al. concluded that there did not appear to be enough
evidence to make alterations of lp based on these tests.
45
Span/depth
L/d
Plastic
hinge
length
lp, in.
(mm)
Ratio to
diameter lp/d
68.01
(469)
4.5
22 (559)
1.38
6.15
(42)
68.01
(469)
4.5
17.05
(433)
1.07
6 (41)
68.01
(469)
4.5
17.01
(432)
1.06
Specimen
Column
diameter
d, in.
(mm)
Concrete
strength
f'c, ksi
(MPa)
Steel
strength
fy, ksi
(MPa)
NF-1
16 (406)
6 (41)
NF-2
16 (406)
9F1
16 (406)
Axial load ratio: P/fcAg 0.3, where P is the column axial load, f c' is the
concrete compressive strength, and Ag is the gross section area of the
column;
eff =
s f ys
(1.28)
f c'
46
Berry et al. (2008) evaluated the plastic hinge length expression provided by
Paulay and Priestley (1992) which has been adopted by California Department of
Transportation. A new expression was proposed by the authors:
l p = 1 D + 2 L + 3
f ydb
(1.29)
f c'
The last term in Eq. 1.29 was added to include the effect of strain penetration on the
spread of plasticity. The new expression was calibrated using the test data and the
parameters (1, 2, and 3) were determined using available optimization algorithms.
Two plastic hinge length expressions were defined as an optimal and a recommended
expression. The optimal expression was determined from the minimum value of the
errors, which were the errors associated with simulation of the column forcedisplacement response history, and the errors due to the ratio of measured-to-calculated
displacements at the beginning of spalling and bar buckling. The optimal expression is
written as:
l p = 0.0375 L + 0.01
f ydb
l p = 0.0375 L + 0.12
f y db
f c'
f c'
47
(psi)
(1.30a)
(MPa)
(1.30b)
l p = 0.05 L + 0.1
f y db
f c'
f ydb
(psi)
(MPa)
f c'
(1.31a)
(1.31b)
The force-displacement response predicted using Eq. (1.31a and b) for columns
with different axial load and aspect ratios shows good comparison with the envelope to
the measured cyclic response, as shown for selected tests in Fig. 1.16.
48
1.4
1.4.1
various aspect ratios and axial-load ratios were tested on a shaking table to investigate
the applicability of current design provisions and theoretical models that are based on
test data from quasi-statically tested specimens, to more realistic dynamic base excitation
conditions that columns experience during earthquakes.
The main parameters of interest were the aspect ratio of the columns, the axial
load ratio, the base flexibility, the base input motions, and the effect of initial low-level
shaking on the response of columns to subsequent higher-level shakes. Table 1.11 lists
details of all 14 columns. All of the columns were 200 mm in diameter and contained 18
49
uniformly distributed longitudinal bars having a 0.24 in. (6 mm) diameter with an
average yield stress of 65 ksi (450 MPa). The longitudinal steel ratio was 1.62%. The
compressive strengths at the time of testing were given in Table 1.11. Columns with
aspect ratios of 4, 7, and 10 were tested to provide data on a range of flexurally
dominated columns. Axial-load ratios of 0.05 and 0.4 were chosen to check the behavior
of columns above and below the balance point on the axial load moment interaction
diagram.
A record based on the 1940 El Centro N-S component earthquake was used as a
base motion for all columns except 1a, 2a, and 3a, which were subjected to a uniform
sinusoidal base motion.
The effective plastic hinge length on a bridge column was considered to be the
equivalent length over which all plasticity takes place with a constant plastic curvature.
The plastic hinge length was calculated using the equation proposed by Priestley et al.
(1992) and the reduction factor defined by Zahn et al. (1986). The reduction factor was
to be 0.5+1.67P/(fcAg) for P/(fcAg) less than 0.3 to modify Priestleys equation. It was
observed from the experimental values that axial load has an effect on the plastic hinge
length and the equation overestimates the plastic hinge length for the columns. It was
also determined from the comparisons of calculated and measured plastic hinge length
values that Zahns recommendation is mostly conservative and agrees well with the
measured values for all the columns having different aspect ratios.
50
fy, ksi
l
(MPa) (%)
(4)
(5)
65
1a
0.04
1.62
(448)
65
1b
0.051
1.62
(448)
65
2a
5 (34) 0.048
1.62
(448)
4.39
65
2b
0.054
1.62
(30)
(448)
5.42
65
3a
0.044
1.62
(37)
(448)
4.22
65
3b
0.056
1.62
(29)
(448)
5.71
65
4
0.391
1.62
(39)
(448)
5.39
65
5
0.388
1.62
(37)
(448)
5.6
65
1.62
(39)
6
0.387 (448)
5.99
65
7
1.62
(41)
(448)
0.04
5.81
65
8
1.62
(40)
0.392 (448)
5.86
65
9
1.62
(40)
(448)
0.041
5.35
65
10
1.62
(37)
0.044 (448)
6.24
65
11
1.62
(43)
0.358 (448)
1: Compressive cylinder strength of concrete
Column
No.
P/fcAg
(3)
Transverse
Steel
fyh, ksi
(MPa)
(6)
37.7
(260)
37.7
(260)
37.7
(260)
37.7
(260)
36.25
(250)
36.25
(250)
37.7
(260)
37.7
(260)
36.25
(250)
36.25
(250)
37.7
(260)
36.25
(250)
37.7
(260)
37.7
(260)
s
(%)
(7)
0.88
0.88
1.06
1.06
0.66
0.66
0.94
1.02
0.79
0.46
0.94
0.46
0.88
0.94
D,
in.
(mm)
(8)
7.9
(201)
7.9
(201)
7.9
(201)
7.9
(201)
7.9
(201)
7.9
(201)
7.9
(201)
7.9
(201)
7.9
(201)
7.9
(201)
7.9
(201)
7.9
(201)
7.9
(201)
7.9
(201)
51
L/D
(9)
4
4
7
7
10
10
4
7
10
4
4
7
4
4
Calc.
Lp,
in.
(mm)
(10)
4.84
(123)
4.84
(123)
6.73
(171)
6.73
(171)
8.62
(219)
8.62
(219)
4.84
(123)
6.73
(171)
8.62
(219)
4.84
(123)
4.84
(123)
6.73
(171)
4.84
(123)
4.84
(123)
Meas.Lp,
in. (mm)
(11)
3.54
(90)
3.35
(85)
4.92
(125)
5.31
(135)
7.68
(195)
5.91
(150)
4.72
(120)
11.02
(280)
10.83
(275)
3.35
(85)
7.48
(190)
5.31
(135)
3.74
(95)
7.68
(195)
1.4.2
52
(305 mm) to 14 in. (356 mm) which was about 0.75D, where D is the diameter of the
columns. The estimated plastic hinge length was within the observed range.
1.5
SUMMARY
Reinforced concrete bridge systems are generally designed to form plastic hinges
in columns rather than the beams in the superstructure. Special detailing must be
provided in reinforced concrete bridge columns along and beyond concentrated yielded
regions (designated as the plastic hinge zones) to provide satisfactory behavior during
response to strong ground motions. In addition, simple models to estimate drift in
reinforced concrete bridge columns can be developed using a definition of lumped
plasticity in the defined plastic hinge zone. Therefore, models that accurately locate the
development and progression of hinging in reinforced concrete bridge systems support
the improvement of calculations for local (damage in hinging regions and beyond) and
global (maximum drift response) performance parameters.
Many researchers have studied the plastic hinge length in column elements
depending on various parameters, such as axial load, yield penetration, shear spreading
of plasticity, shear span-to-depth ratio, and the amount of longitudinal reinforcement.
Various expressions were proposed by several researchers to estimate the plastic hinge
length in these elements. The following are the observations investigated by the previous
researchers:
Axial Load. Extensive research has been conducted to examine the influence
the level of axial load on plastic hinge length (ICC Committee Report (1962),
Park et al. (1982), Priestley and Park (1987), Sakai and Sheikh (1989),
Tanaka and Park (1990), Sheikh (1993, and 1994), Watson and Park (1994),
Kovacic (1995), Bayrak and Sheikh (1997, and 1999), Dodd et al. (2000), and
53
lc
P
= 1 + 0.4 '
h
f c Ag
P
= 0.3
h P0
lp
As
+ 3
Ag
(Bae 2005)
(1.18a)
(1.20)
Yield Penetration and Shear Spreading. Mander (1983), Priestley and Park
(1987), Paulay and Priestley (1992), and Restrepo (2006) investigated the
effects of yield penetration and shear spreading of plasticity. They proposed
their formulae with respect to the longitudinal reinforcement diameter (db)
and the column height (L or H) based on the experimental results. Paulay and
Priestley (1992), and Restrepo (2006) also considered the effect of yield
stress of the longitudinal reinforcement in their expressions:
54
l p = 0.08 L + 6d b
(1.16)
(1.17a)
L p = H + f ye d bl 2 f ye d bl
fle
= 1 1 2
=
f
f ye d bl
M ' H2
(
y )
3
My
pH 2
L p = L py +0.06 L
L py = 6.35 d b
l p = 0.05 L + 0.008
f y db
f c'
(Restrepo 2006)
(1.21)
(Mander 1983)
(1.13)
(1.31a)
Shear Span-to-Depth Ratio. Shear span-to-depth ratio was one of the main
variables for Sakai and Sheikh (1989), Tanaka and Park (1990), Bayrak and
Sheikh (1997, and 1999) and Bae (2005). All of the observations given by the
researchers above were consistent that plastic hinge length increased with
increasing shear span-to-depth ratio. Bae (2005) first attempted to study the
relationship between the plastic hinge length and the shear span-to-depth ratio
and proposed a new expression:
P
= 0.3
h P0
lp
As
+ 3
Ag
(Bae 2005)
(1.20)
Member Dimensions. Early expressions for plastic hinge length used the
relationship between the column height, the cross sectional depth, and/or the
diameter of longitudinal reinforcement with the plastic hinge length. The
column height (L) and the depth (d or h) represented the flexural response of
55
the
columns,
whereas
the
diameter
of
longitudinal
reinforcement
Dynamic Tests. The studies conducted by Dodd et al. (2000), and Hachem et
al. (2003) utilized the equation proposed by Priestley et al. (1992) to estimate
the plastic hinge length before testing the circular columns under earthquake
loading . This equation was developed using the static test results of 20 bridge
columns having a shear-span-to-depth ratio between 2 and 5.5. According to
dynamic test results, it was concluded that the estimated plastic hinge length
values based on static tests overestimated the measured plastic hinge length
for some columns, but were within a reasonable range of the measured
dynamic test results for other columns. This deserves further study, and will
be investigated in this research.
The expressions that have been developed in the literature are based on
component tests and analyses of their data. The question that remains is how these
estimations of plastic hinge length, damage locations, and estimates of column drift may
change in large scale system tests. This question is the focus of the current study.
56
1.6
The objective of the current research is to study the plastic hinge length, the
effect of the spread of plasticity on estimates of drift, and the progression of damage in
reinforced concrete bridge systems during response to earthquake motions. The scope of
research to accomplish these objectives includes the followings:
Based on this analysis, a new expression is proposed for plastic hinge length.
Finally, an analytical approach for the spread of plasticity and simple drift
calculations are also developed.
57
2.1
INTRODUCTION
photogrammetry method was used to track deformations in the hinging regions of one
bridge column in a large-scale four-span bridge test. Both tests employed strong ground
motion to excite the structure, with the first test in one direction and the second with bidirectional loading.
58
2.2
PROOF-OF-CONCEPT TEST
On December 14, 2005, a proof-of-concept test for photogrammetry
59
Spiral Reinf
0.25 in.
clearcover=0.5 in. [13 mm]
Fig. 2.1 Cross-section of the column
In order to keep track of the displacements on the column surface during the test,
a grid system was applied to the column surface (Fig. 2.2). Figure 2.2 also shows the
traditional linear vertical displacement transducers (LVDT) instrument that was used
during the test. The grid was applied to the surface using a cardboard stencil at the
University of Kansas and black and white paint. First a uniform black background was
applied, and then the stencil was used to apply the white grid marks. The grid consisted
of vertical and horizontal lines spaced irregularly around the face of the circular column.
The irregular spacing of the grid marks allows for a detailed analysis of the column
circular surface. The grid system was applied above the column plastic hinging region to
avoid the congestion of LVDT instrumentation. Displacement and rotation calculations
are reported for the location marked A in Fig. 2.2.
60
Load Direction
Mass
Rig
A system of two cameras: model name DXB-9212EF and 0.5-in. (13 mm)
Starlight 600 TVL super high resolution black-white recording properties.
The angles from left and right cameras to the column were 60 (Fig. 2.3).
VCR systems: JVC HR-S2902U brand and high resolution stereo video
cassette recorder.
Monitor: a black and white monitor to show the field of view of the
camera.
An aluminum tower:
cameras. Only the first level of the tower used in the proof-of-concept
test. Because the height of the first level was 94.5 in. (2400 mm), metal
and masonry units were placed under the legs of the tower to enable the
cameras to photograph at the level of the grid (Fig. 2.4).
Connections: small aluminum connections which are 30, 45 and 60angle pieces.
86
3.
10
m
]
]3
6
in
.
60
RC
67
.8
6
in
.[
1.
72
m]
60
SVTN
n.
[2.
24
88
i
60
64
2.
.[
LC
[0
.9
1
m]
in
88
[
in.
4
2.2
Shake Table
62
Tower
63
The
Acceleration (g)
0.4
0.2
0
0
10
15
-0.2
-0.4
-0.6
-0.8
Time (sec.)
Due to setup limitations in the lab, the distance between the cameras and the
column was 88 in. At this distance, the field of view was too narrow to include two
vertical lines in the photographs at all times. Therefore, the proof-of-concept test focused
on vertical rotation of the column and did not consider the cross-sectional rotation of the
column.
Calculated
displacements
and
rotations
using
the
LVDT
data
and
photogrammetry measurements are compared at Point A on the painted grid (Fig. 2.2).
Figure 2.6 shows the comparison for calculated displacements. Because of a limitation
of the number of frames that may be considered in a single analysis, the photogrammetry
results are shown in two pieces (Run1L and Run2L). The results from two runs are
superimposed and demonstrate good continuity for calculated displacements.
The LVDT displacement values are obtained by first considering the relative
displacement between the displacement transducer recordings at the top of the column
and the absolute displacement of the shaking table.
64
estimated by linearly interpolating between the relative displacement at the top of the
column and zero displacement at the base to the height of Point A. Displacements values
using the photogrammetry data are determined as the relative motion of Point A. As
shown in Fig. 2.6, the displacement results in terms of periodicity and amplitude of
motion are well-represented by the photogrammetry analysis. The permanent
deformation noted at the end of the test is also captured in the photogrammetry analysis.
Rotations of the horizontal line are compared in Fig. 2.7.
Rotations of the
horizontal line Cross section rotations using the LVDT data are first estimated at each
LVDT location by dividing the relative vertical displacement between two opposing
LVDTs by the column width. As expected, the change in cross section rotation over the
column height of the hinging region decreased with increasing column height.
To
estimate the rotation at Point A, it was assumed that the change in rotation between the
two highest LVDT locations remained constant to Point A.
Rotations of the cross-section based on photogrammetry measurements were
estimated assuming no shear distortions occurred in the grid region.
Using this
65
3.0
2.5
2.0
X-Displacement (in.)
1.5
1.0
0.5
0.0
-0.5
-1.0
-1.5
-2.0
-2.5
-3.0
0
10
11
12
13
14
15
16
17
18
19
20
Time (sec.)
LVDT
Run 1L
Run 2L
Fig. 2.6 Displacement @ Grid Level A vs. Time (Rinaldi 0.95g) (1 in. = 254 mm)
4.0
3.5
3.0
2.5
Rotation (degrees)
2.0
1.5
1.0
0.5
0.0
-0.5
-1.0
-1.5
-2.0
-2.5
0
10
11
12
13
14
15
16
Time (sec.)
LVDT
Run 1L
Run 2L
66
17
18
19
20
2.3
2007 at the UNR laboratory. Representatives from NEES, CALTRANS, UNR, KU,
other participating universities, and local media congregated in the test area.
Photogrammetry measurements were taken in two hinging regions, which required the
use of four cameras and two grid configurations. The significance of this test was that it
was a large-scale bridge with columns having double curvature response and resisting
massive earthquake forces. Also, biaxial earthquake loading was employed. Challenges
that are addressed in the photogrammetry analysis included irregular image quality and
inconsistent definition of points on the concrete surface.
2.3.1
Description of Specimen
The test specimen consisted of a four-span reinforced concrete bridge with end
abutments as shown in (Fig. 2.8). The bridge was quarter-scale with two interior spans
that were 29 ft. (8.84 m) in length, and two exterior spans of 24.5 ft. (7.47 m) for a total
length of approximately 110 ft. (33.5 m). The clear heights of the bents were 5, 6 and 7
ft. (1.52, 1.83, and 2.13 m), with the tallest bent in the middle (Bent 2). The
superstructure consisted of a solid slab that was post-tensioned in both the longitudinal
and transverse directions. The bridge also included abutment seats at both ends that were
driven in the longitudinal direction by dynamic hydraulic actuators. The depth of the cap
beams was 15 in. (0.38 m). The total length of the cap beams (perpendicular to the
longitudinal axis of the bridge) was 98 in. (2.49 m). The concrete compressive strength
used in the bridge (excluding the slab) was 6.7 ksi (46 MPa) and reinforcement was
ASTM A706 Grade 60. The bents were numbered Bent 1, Bent 2, and Bent 3 starting
from left (South) to right (North). The east column of Bent 3 was selected to be the target
67
structure for applying the painted grids. Figures 2.9, 2.10 and 2.11 show the elevation
and side views of Bents 1, 2 and 3, respectively.
N
SHAKE TABLE
Abutment 1
SHAKE TABLE
Bent 1
SHAKE TABLE
Bent 2
Bent 3
Abutment 2
15in [0.38m]
98in [2.49m]
60in [1.52m]
1.5in [0.04m]
5.5in [140mm]
63in [1.6m]
12in [305mm]
12in [305mm]
5.5in [140mm]
15in [381mm]
12in [305mm]
18in [457mm]
36in [0.91m]
108in [2.74m]
68
1 5 in [ 3 8 1 m m ]
9 8 in [2 .4 9 m ]
6 3 in [1 .6 m ]
1.5in [38mm]
5.5in [140mm]
12in [305mm]
12in [305mm]
5.5in [140mm]
1 5 in [3 8 1 m m ]
1 2 in [3 0 5 m m ]
8 4 in [ 2 . 1 3 m ]
1 8 in [4 5 7 m m ]
1 0 8 in [2 .7 4 m ]
3 6 in [ 0 . 9 1 m ]
98in [2.49m]
1 5in [0 .3 8m ]
1.5in [0.04m]
5.5in [140mm]
63in [1.6m]
12in [305mm]
12in [305mm]
5.5in [140mm]
15in [381mm]
7 2in [1.83 m ]
1 2in [30 5m m ]
3 0in [0 .7 6m ]
36 in [0.91 m ]
108in [2.74m]
2.3.2
Experimental Setup
The top and bottom hinging regions of the east column of Bent 3 were the focus
concept test. The top grid used a different configuration consisting of lines and squares
(Fig. 2.3b) that resembled the spacing of the lines in the bottom grid. The top grid was
applied to track the displacements of the isolated points (squares), similar to applications
in earlier photogrammetry methods that were completed at the University of Kansas
laboratory (Reetz, 2005).
To apply the grid systems, the column was first spray-painted black for a solid
dark background. The lines and squares were lightly traced with a pencil, and then a
single stencil was used to fill in the lines and squares using a white paint.
The same aluminum tower as described in the proof-of-concept test Section (2.2)
was used to support the cameras, including the second level of the tower for the tests.
Two cameras were mounted on each level (at 94.5 in. (2400 mm) and 159 in. (4039 mm)
height) (Fig. 2.14). In the first level of the tower, due to changes in the height level of
the hinge region, the height of the cameras was increased to directly view the bottom
grid region. This was accomplished by placing small aluminum columns to mount the
cameras on the tower (Fig. 2.14). Metal weight pieces were placed at each side of the
base of the tower to provide balance on the strong floor (Fig. 2.15). There were four
VCR systems and two monitors (two channels for each monitor) used in the test.
70
(a)
(b)
Fig. 2.13 Close view of (a) bottom and (b) top grid systems
71
Fig. 2.15 Metal weight pieces placed on each side of the tower
72
The distance between the east column of Bent 3 and the tower was 288 in. (7315
mm). The angle from the left camera to the column was 73.58, and from the right
camera to the column was 44.97 on the horizontal plane (Fig. 2.16). The tower was
located on the north-east side of the UNR laboratory.
Figure 2.17 shows the numbers of all points defined by the grid systems. The
points on the bottom and top grid systems were defined at the intersections of the thick
vertical and horizontal lines. The locations of a series of linear vertical displacement
transducers (LVDTs) and one displacement transducer (DT7) monitored by the research
team at UNR are also shown in Fig.2.17.
m]
6.
75.24
14
n.
8i
.9 1
50 ]
3m
3 .7
in .
9
1 .2
m]
6 7 .8
RC
75.
6 in .
[ 1 .7 2
m]
24
Tower
288
LC
1 0 3 .8
6 in.
[2.64
m]
3 6 in
. [0 .9
1
.32m
in. [7
28
8
.[
7.
32
.5 3
in
29
m
]
Bent 3
Shake Table
LC Location of Left Camera
RC Location of Right Camera
Fig. 2.16 Location of the Bent 3 east column and the aluminum tower
73
DT7
3ETR3
57
56 58
50
49 51
59
52
62
6061
55
5354
43
42 44
45
48
4647
3ETR5
3ETR6
36
35 37
29
28 30
212223
41
3940
34
31 3233
24 252627
38
1617 18
19 20
1112
13
1415
6 7
9 10
1 2
4 5
3EBR5
3EBR7
3ETR4
3EBR6
3EBR8
Fig. 2.17 Grid system and LVDT locations on column in the Bent 3 east column
2.3.3
Earthquake Loading
The earthquake motion used in the tests consisted of biaxial and uniaxial
applications of scaled motions measured at the Century City Country Club during the
1994 Northridge, California earthquake. This motion was applied 13 times with six
different scaled intensities having increased amplitude so that the progression of damage
could be tracked from pre-yield to failure as shown in Table 2.1.
A white noise
excitation was applied before each change in ground motion intensity to test for system
identification.
The first damage to the bridge system occurred in the columns of Bent 3 (the bent
with medium height) and then the columns of Bent 1 (the bent with the shortest height).
74
The earthquake intensity was not increased after the 1.20g motion (Test 6) because the
reinforcement in the left column of Bent 1 started to buckle.
longitudinal bars in the columns of Bent 1 had buckled by Test 6. The same motion was
repeated one more time before the test was completed.
Other damage that was noted during the test included small amounts of cracking
in the bent caps, and some spalling of the cover concrete of the abutment seats and the
superstructure. The cracks at the bent caps began after the first test (Test 1A) and
progressed until the tests were done. The spalling of the concrete cover of the abutment
seats occurred after Test 6.
75
2.3.4
Test Date
2/12/2007
Motion
Level
1
1
1
1
2/13/2007
4
4
4
4
2/14/2007
5
2/15/2007
Results
Due to shaking effects on the strong floor during the test, there is a chance for
movement of the base of the aluminum tower. In order to eliminate the extra movement
76
from the tower shakings, a fixed point on the wall was identified and tracked during the
strong excitations (Fig. 2.18).
Fig. 2.18 Fix point on the wall at the back of the column
The lateral displacement of the fixed point on the wall (Fig. 2.19) was determined
and subtracted from the lateral displacements of the grid points to obtain the
displacements of the column caused by the shake table motions. The calculated tower
displacements as indicated by the movement of the fix point are very close to zero in
value (within one or two pixels of value) and did not significantly affect the results. The
vertical displacement of the fixed point was also calculated (Fig. 2.20), but it does not
need to be subtracted from the vertical displacement of the grid point because crosssectional rotation calculations are based on relative vertical movement between two
points on the column.
77
0.1
0.05
0
0
10
15
20
25
30
35
-0.05
-0.1
Time (sec.)
Fig. 2.19 Lateral displacement of fixed point on the wall at Test 4D (1 in. = 254 mm)
0.015
0.01
0.005
0
-0.005 0
10
15
20
25
30
35
-0.01
-0.015
Time (sec.)
Fig. 2.20 Vertical displacement of fixed point on the wall at Test 4D (1 in. = 254 mm)
There are some differences in the data provided by LVDT instrumentation and
that recorded for photogrammetry application. The rotation calculations obtained using
photogrammetry analysis represents the cross-sectional rotations based on the movement
of the column, which includes not only flexural but also rigid body rotation. Rotations
calculated using the LVDT data from the test includes only flexural contributions. The
LVDT instrumentation was set to record movement in the transverse and longitudinal
planes, and photogrammetry measuruments were recorded at an angle between the
LVDT instrumentation. This difference in orientation is accounted for in the analysis
and results presented in this section.
Unclear Images
Interactive Data Language (IDL) was used to analyze the video images that were
recorded during the test. Each video was first divided into over one thousand images for
individual analysis. The displacement of a unique point in an image was defined as the
relative horizontal movement of that point between successive images. However, at
times it was difficult to keep track of a unique point because of the quality of the images.
The image (640 x 480) was written as 480 lines that are 640 pixels wide. The
odd-numbered lines are written first, followed by the even-numbered lines so that the
image is really two images offset by 1/60th of a second. The first partial image (half of
the total image) is comprised of the odd lines, and the second partial image is
comprised of the even lines. Due to the shaking effect from the strong floor during the
tests and the incompatibility in recording frequency between the cameras and the
earthquake motions, many images were compromised. Some of the images appeared as
double lines for each horizontal and vertical line, and some were simply blurry images.
Table 2.2 shows the number of corrupted images that were recorded during each test
79
motion and the total percentage of corrupted images recorded by each camera. The
Image Group number in the table signifies a grouping of images that could be analyzed
at one time. Therefore, Image Group 1 contains images from the earliest time in a test.
Each image represents 1/30th of a second during the test. The total number of images
reviewed in a single analysis ranged from 339 to 998. The percentage of corrupted
images ranges from 0-62%. For recordings at the top of the column, the percentage of
corrupted images exceeded 4% when at least one component of shaking exceeded 0.5g
with a maximum of 30% corrupted images (Test 4D Top Right Camera). Recording
quality was much more sensitive at the base of the column, with as much as 24%
corrupted images during Test 2 (Bottom Left Camera), which had a peak ground
acceleration of only 0.15g. The images recorded early (Image Group 1) in each shake
table test ranged from 8-62% corrupted, the recordings from Test 5 Bottom Left Camera
having 62% corrupted images.
80
Top Left
2
Top Right
2
Bottom Left
2
Bottom Right
2
Image
Groups
No. of
corrupted
images
27
28
80
72
No. of total
images
899
299
899
349
339
339
339
339
339
339
Corruption
Percentage
24
21
Test
4D
4D
4D
4D
Image
Groups
No. of
corrupted
images
100
102
10
171
40
193
10
No. of total
images
339
339
339
339
339
339
339
339
339
339
339
339
Corruption
Percentage
29
30
50
12
57
Test
Image
Groups
No. of
corrupted
images
220
136
210
55
180
15
No. of total
images
899
899
339
339
339
339
339 339
Corruption
Percentage
24
15
62
16
53
Test
Image
Groups
No. of
corrupted
images
210
200
221
264
No. of total
images
999
339
999
339
999
339
999
339
Corruption
Percentage
21
20
22
26
81
Even
Odd
0.20
Comb.
0.15
0.10
0.05
0.00
-0.05
10
15
20
25
30
-0.10
-0.15
-0.20
Time (sec.)
Fig. 2.21 Point 7 vertical displacement at the bottom grid system (1 in. = 254 mm)
82
35
Even
Vertical Displacement (in.)
0.20
0.15
0.10
0.05
0.00
5
10
-0.05
-0.10
-0.15
-0.20
Time (sec.)
a) Even Line
Point 7 Vertical Displacement (Test4D Bottom)
0.25
Odd
Vertical Displacement (in.)
0.20
0.15
0.10
0.05
0.00
5
10
-0.05
-0.10
-0.15
-0.20
Time (sec.)
b) Odd Line
Point 7 Vertical Displacement (Test4D Bottom)
0.25
Comb.
0.20
0.15
0.10
0.05
0.00
5
10
-0.05
-0.10
-0.15
-0.20
Time (sec.)
c) Combination
Fig. 2.22 Five second interval to compare even- odd- and combination-line analyses (1
in. = 254 mm)
83
The vertical displacement results of Point 7 in the even and odd line analyses are
compared with the LVDT 3EBR7 reading (the closest UNR measurement to the point on
the bottom grid system as shown in Fig. 2.17) in Figs. 2.23-2.24. Figure 2.25 shows only
the LVDT 3EBR7 recorded displacements for clarity. Because of the larger disparity
noted in the figures between peak displacement values recorded in the odd-line analysis
(Fig. 2.24) and LVDT recordings, the even line approach (Fig. 2.23) is selected to
provide better photogrammetry results.
Rotation of horizontal line is calculated as the relative vertical displacements of
the horizontal line that Points 7 and 8 align and divided by the horizontal distance
between the points. For the photogrammetry analysis, the horizontal distance between
the points changes within each image. Figure 2.26 shows the rotation of horizontal line
calculated using Points 7 and 8 at 12 inches (305 mm) from the bottom fixity in the
transverse direction. The appearance of noise with nearly constant magnitude is a
result of single pixel differences in calculated displacements (pixelization).
Point 7 Vertical Displacement (Test4D Bottom)
0.25
Even
LVDT-3EBR7
0.20
0.15
0.10
0.05
0.00
-0.05
10
15
20
25
30
35
-0.10
-0.15
-0.20
Time (sec.)
Fig. 2.23 Comparison of Point 7 vertical displacement for even lines with LVDT 3EBR7
data (1 in. = 254 mm)
84
0.20
LVDT-3EBR7
0.15
0.10
0.05
0.00
0
10
15
20
25
30
35
-0.05
-0.10
-0.15
-0.20
Time (sec.)
Fig. 2.24 Comparison of Point 7 vertical displacement for odd lines with LVDT 3EBR7
data (1 in. = 254 mm)
0.25
0.20
0.15
0.10
0.05
0.00
-0.05 0
-0.10
10
15
20
25
30
-0.15
-0.20
Time (sec.)
Fig. 2.25 LVDT 3EBR7 vertical displacement history (1 in. = 254 mm)
85
35
0.030
0.025
0.020
0.015
Rotation (rad)
0.010
0.005
0.000
0
10
15
20
25
30
35
-0.005
-0.010
-0.015
-0.020
-0.025
Time (sec.)
Fig. 2.26 Rotation of horizontal line calculated from Points 7 and 8 at the bottom grid in
the transverse direction
There are several options for defining each unique point on the grid, all of which
were considered and the best overall representation selected as described in this section.
Points on the grid are defined by the intersection of the thick vertical and horizontal
lines. The vertical and horizontal lines are 0.25 in. (6 mm) thick; therefore, there are
several ways to define a line within that thickness. One option is to seek a definition of
the mid-point of each 0.25-in. (6 mm) thickness and define a middle line as the
connection of those points (Fig. 2.27). Another option is to define the edges of the lines.
For this option, the edge lines may be defined by points of a particular light intensity
value, or a change in relative intensity between the black to white paint transition.
86
Finally, the length of the line that is used to define each Point on the grid also will
influence results. Each of these options is discussed below.
Middle Line
Edge Lines
Edge Line Point 3
Middle Line Point 3
Fig. 2.27 Close up of lines used to define Point 3 (Fig. 2.17)
1. Middle Lines: In IDL program, the intensity of each pixel was determined
for each image in the analysis. To define the middle lines, a threshold value
of light intensity (Robert threshold) was used to define the left and right
boundaries for vertical lines and top and bottom boundaries for horizontal
lines in the images. The definitions of the boundaries of the lines were
completed for a cropped image area. The intensity of each pixel of a given
line was then determined between the defined boundaries for the vertical and
horizontal lines. In theory, the maximum pixel intensity would define a
unique line through the middle portion of each 0.25-in. (6 mm) line. The
locations of the maximum pixel value along the thickness of the lines near
Point 8 are shown in Fig. 2.28. To get refined results for the maximum
intensity pixel locations, the area around the maximum intensity pixel
locations were searched in detail. Because the location of the maximum
pixel intensity varied over the length of each line, the middle points did not
87
align to form a smooth middle line. Therefore, the displacements and the
cross-sectional rotation results obtained using this method were not very
accurate.
The vertical displacements of Point 7 and 9 are compared with LVDT
3EBR7 and 3EBR8 readings in Figs. 2.29 and 2.30.
As shown in the
figures, the peak values represent LVDT measurement fairly well, but there
is a lot of high frequency noise in the plots. The resulting calculated rotation
at about 12 in. (305 mm) from the footing of Bent3 is shown in Fig. 2.31.
As expected, the detailed calculation of the rotation varies widely.
225
133 180 214
219
102 149 189 214
88
0.25
0.20
0.15
0.10
0.05
0.00
0
10
15
20
25
30
35
-0.05
-0.10
-0.15
-0.20
Time (sec.)
Fig. 2.29 Comparison of Point 7 vertical displacement with LVDT 3EBR7 data (1 in. =
254 mm)
Point 9 Vertical Displacement (Test 4D Bottom)
0.30
Middle Line
LVDT-3EBR8
0.25
0.20
0.15
0.10
0.05
0.00
0
10
15
20
25
30
35
-0.05
-0.10
-0.15
-0.20
Time (sec.)
Fig. 2.30 Comparison of Point 9 vertical displacement with LVDT 3EBR8 data (1 in. =
254 mm)
89
0.030
0.025
0.020
Rotation (rad)
0.015
0.010
0.005
0.000
0
10
15
20
25
30
35
-0.005
-0.010
Time (sec.)
90
0.15
IDL_h=7.6 - 13.2 in.
Rotation (rad)
0.10
0.05
0.00
0
10
15
20
25
30
35
40
-0.05
Time (sec.)
2. Edge Lines: Because of the thickness of the horizontal and vertical lines that
define the grid system, the exact unique location of each Point (generally
located at the intersection of the thick horizontal and vertical grid lines) may
be defined in several ways.
previously, did not yield reliable results. Another method is to consider the
edges of the thick grid lines to define two new lines (edge-lines) for each
grid line. Therefore, at the intersection of the thick grid lines that generally
defines a Point, there are actually four intersections of the edge lines to
consider as possible unique locations of that Point (see Fig. 2.33, for
example, Point 12 and Point 13 a, b, c, and d).
In IDL programming
language, the equations of the edge lines were obtained by defining the
91
boundaries of the thick grid lines using the constant Robert threshold value,
as described in the previous section.
Figures 2.34 through 2.37 show the rotation of horizontal line results
calculated between the corners of the Points.
rotations were averaged to obtain the middle point (Point 12 and 13)
rotation response. Figure 2.38 shows the middle point rotation result for
Test 4D. The middle Point was selected to calculate the rotations for
photogrammetry results because the corner rotation results give high values
and more high frequency noise than the middle point rotation result.
13
12
c
92
0.15
Rotation (rad)
0.10
0.05
0.00
0
10
15
20
25
30
35
30
35
-0.05
-0.10
Time (sec.)
0.15
Rotation (rad)
0.10
0.05
0.00
0
10
15
20
25
-0.05
-0.10
Time (sec.)
93
0.15
Rotation (rad)
0.10
0.05
0.00
0
10
15
20
25
30
35
30
35
-0.05
-0.10
Time (sec.)
0.15
Rotation (rad)
0.10
0.05
0.00
0
10
15
20
25
-0.05
-0.10
Time (sec.)
94
0.08
Rotation (rad)
0.06
0.04
0.02
0.00
0
10
15
20
25
30
35
-0.02
-0.04
Time (sec.)
threshold value that adequately defined the transition from black to white
paint was determined from the first image in each image group. A line
length was assigned and the equations of the four edge lines were calculated.
Various degrees of polynomials were considered to define the edge lines, but
simple first degree (linear) representation proved to be the best fit. The edge
line lengths were set between values of 0.5 in. (13 mm) and 1.0 in. (25 mm)
depending on the relative distance between the grid lines. Even-numbered
lines were used in the analysis, as described previously (Section 2.3.4.1
Unclear Images).
95
96
Robert Threshold
LVDT_3EBR7
0.20
0.15
0.10
0.05
0.00
0
10
15
20
25
30
35
-0.05
-0.10
-0.15
Time (sec.)
Fig. 2.39 Point 7 vertical displacement vs. LVDT 3EBR7 data (1 in. = 254 mm)
V e rtic a l D is p la c e m e n t (in .)
0.25
Robert Threshold
LVDT-3EBR8
0.20
0.15
0.10
0.05
0.00
0
10
15
20
25
30
35
-0.05
-0.10
-0.15
Time (sec.)
Fig. 2.40 Point 9 vertical displacement vs. LVDT 3EBR8 data (1 in. = 254 mm)
97
0.030
0.025
0.020
Rotation (rad)
0.015
0.010
0.005
0.000
0
10
15
20
25
30
35
-0.005
-0.010
Time (sec.)
Fig. 2.41 Rotation of horizontal line obtained using the constant Robert threshold
Point 7 Vertical Displacement (Test 4D Bottom)
0.30
Intensity Difference
0.25
LVDT-3EBR7
0.20
0.15
0.10
0.05
0.00
-0.05
10
15
20
25
30
35
-0.10
-0.15
-0.20
Time (sec.)
Fig. 2.42 Point 7 vertical displacement compared with LVDT 3EBR7 data (1 in. = 254
mm)
98
LVDT-3EBR8
0.20
0.10
0.00
0
10
15
20
25
30
35
-0.10
-0.20
Time (sec.)
Fig. 2.43 Point 9 vertical displacement compared with LVDT 3EBR8 data (1 in. = 254
mm)
0.06
Rotation (rad)
0.04
0.02
0.00
0
10
15
20
25
30
-0.02
-0.04
-0.06
Time (sec.)
Fig. 2.44 Rotation of horizontal line at h=7.7 in. (196 mm) from the bottom fixity
99
35
4. Long Lines vs. Short Lines: When using the constant Robert
threshold method, the length of the edge line controls the number of points
used to define that line, and therefore, the equation of the line. Longer (1.5
in. (38 mm) to 3.0 in. (76 mm)) edge-line lengths were investigated to
determine the influence on the definition of a unique Point. Even-numbered
lines were used in the analysis.
The vertical displacements of Point 7 and 9 were determined and
compared with LVDT 3EBR7 and 3EBR8 readings (Figs. 2.45-2.46). The
resulting cross-sectional rotation is shown in Fig. 2.47 and compared with
the calculated rotation from LVDT data at approximately 12 in. from the
footing of Bent3. The short line and long line analyses can be compared by
examining cross-sectional rotation calculated from each analysis (Figs. 2.41
and 2.47, respectively). There is approximately 0.011 radians of constant
noise shown in the long-line results (Fig. 2.47), whereas the short-line
results (Fig. 2.41) show approximately half the magnitude and less
frequency of noise. The short-line definition was selected for all ensuing
calculations.
100
0.20
LVDT-3EBR7
0.15
0.10
0.05
0.00
0
10
15
20
25
30
35
-0.05
-0.10
-0.15
Time (sec.)
Fig. 2.45 Comparison of Point 7 vertical displacement with LVDT 3EBR7 data (1 in. =
254 mm)
Point 9 Vertical Displacement (Test 4D Bottom)
0.25
Long-Line
V e rtic a l D is p la c e m e n t (in .)
0.20
LVDT-3EBR8
0.15
0.10
0.05
0.00
0
10
15
20
25
30
35
-0.05
-0.10
-0.15
Time (sec.)
Fig. 2.46 Comparison of Point 9 vertical displacement with LVDT 3EBR8 data (1 in. =
254 mm)
101
0.030
0.025
0.020
Rotation (rad)
0.015
0.010
0.005
0.000
0
10
15
20
25
30
35
-0.005
-0.010
Time (sec.)
102
calculate corner points and the Average Point location is shown in Fig. 2.49 and
compared with LVDT data. As shown in both figures, the vertical displacement
of the middle point follows the LVDT 3ETR4 data when the Edge Lines method
is used and gives much closer results than the vertical displacement using the
maximum intensity approach.
LVDT-3ETR4
0.25
0.20
0.15
0.10
0.05
0.00
0
10
15
20
25
30
35
-0.05
-0.10
-0.15
-0.20
-0.25
-0.30
Time (sec.)
Fig. 2.48 Comparison of Point 46 vertical displacement with LVDT 3ETR4 data
using the maximum intensity approach (1 in. = 254 mm)
103
LVDT-3ETR4
0.25
0.20
0.15
0.10
0.05
0.00
0
10
15
20
25
30
35
-0.05
-0.10
-0.15
-0.20
-0.25
-0.30
Time (sec.)
Fig. 2.49 Comparison of Point 46 vertical displacement with LVDT 3ETR4 data using
the Edge Lines method (1 in. = 254 mm)
104
The lateral displacements of Point 3 and 8 on the bottom grid were tracked during
Test 4D motion to calculate rotations of vertical line (Fig. 2.50). Points 7 and 9 shown in
Fig. 2.17 were selected and the vertical movements of the Points were determined to
calculate the rotation of horizontal line during Test 4D. The rotation of horizontal line
results is shown in Fig. 2.51. Also, the rotation of vertical line and the rotation of
horizontal line are compared in Fig. 2.52.
In order to compare the rotation of vertical line and rotation of horizontal line at
the top grid, Point 38 and 52 were selected and tracked during Test 4D. Figure 2.53
shows the rotation of vertical line results calculated between Point 38 and 52. Point 44
and 46 were selected to determine the rotation of horizontal line shown in Fig. 2.54. The
comparison between the rotation of vertical line and rotation of horizontal line results are
shown in Fig. 2.55. As an average measurement, the rotation of vertical line provides a
smoother trace than the rotation of horizontal line calculations from the photogrammetry
data. Comparisons for the three different tests at each cross-sectional rotation location
are found in Appendix A.
105
0.02
0.015
Rotation (rad)
0.01
0.005
0
0
10
15
20
25
30
35
-0.005
-0.01
Time (sec.)
0.02
0.015
0.01
Rotation (rad)
0.005
0
-0.005
10
15
20
25
-0.01
-0.015
-0.02
-0.025
-0.03
Time (sec.)
106
30
35
0.02
Vertical_Rotation
0.015
Cross-sec. Rot.
0.01
Rotation (rad)
0.005
0
0
10
15
20
25
30
35
-0.005
-0.01
-0.015
-0.02
-0.025
-0.03
Time (sec.)
0.04
0.03
Rotation (rad)
0.02
0.01
0
0
10
15
20
25
30
-0.01
-0.02
-0.03
Time (sec.)
35
0.05
0.04
Rotation (rad)
0.03
0.02
0.01
0
0
10
15
20
25
30
35
-0.01
-0.02
-0.03
Time (sec.)
0.05
Vertical_Rotation
0.04
Cross-sec. Rot.
0.03
Rotation (rad)
0.02
0.01
0
0
10
15
20
25
30
-0.01
-0.02
-0.03
Time (sec.)
108
35
2.3.4.2 Displacements
The grid that is used to provide photogrammetry data is advantageous when
considering the deformation to the column surface over a continuum, which may be
correlated with detailed Finite Element results. By contrast, LVDT instrumentation will
provide only discrete displacement data in a single direction. In addition, if significant
damage occurs to the concrete surface, LVDTs measurements may become
dysfunctional, but the painted grid surface may be reconstructed in an image and
photogrammetry data preserved. This section describes the displacement calculations in
the gridded surfaces for multiple earthquake excitations. Displacement histories for all
gridded Points and earthquake motions are shown in Appendix B.
Point 59 in the top grid system (Fig. 2.17), which was the closest point to the top
of the column, was selected to track the peak displacement along the direction that the
left cameras have 15.25 counter clockwise from the longitudinal direction of the bridge.
The displacements at grid Point 59 captured during Test 2, Test 4D, and Test 6 are
compared with the combined components of DT7, which recorded the transverse
displacement of the Bent 3 at the middle of the bridge deck, and DS5 and DS1
measurements, which measured the gap size history and the displacement of the north
abutment in the longitudinal direction of the bridge, during the same motions in Figs.
2.56, 2.57, and 2.58, respectively. There is approximately 23.3 in. (592 mm) between
the transverse beam and Point 59. Because the column is in double curvature during the
test, the lateral displacements at these two heights are assumed to be similar.
109
displacement results and the readings of the displacement transducers are noted (for
example, between the time periods of 12 and 13.5 seconds in Fig. 2.56), and in these
cases the photogrammetry results have been visually confirmed with the associated
images. As shown in all figures, the lateral displacements are reliable and correlate with
the LVDT instrumentation data.
0.8
Photogrammetry
0.6
0.4
Displacement (in.)
0.2
0
0
10
15
20
25
30
35
-0.2
-0.4
-0.6
-0.8
-1
Time (sec.)
Fig. 2.56 Comparison of Drift at Point 59 for Test 2 (1 in. = 254 mm)
110
40
2.5
Photogrammetry
DT7 & (DS5+DS1) combined
components
Displacement (in.)
1.5
0.5
0
0
10
15
20
25
30
35
40
-0.5
-1
-1.5
Time (sec.)
Fig. 2.57 Comparison of Drift at Point 59 for Test 4D (1 in. = 254 mm)
5.0
Photogrammetry
4.0
Displacement (in.)
3.0
2.0
1.0
0.0
0
10
15
20
25
30
35
-1.0
-2.0
-3.0
Time (sec.)
Fig. 2.58 Comparison of Drift at Point 59 for Test 6 (1 in. = 254 mm)
111
40
Figure 2.59
shows the ratios for the three different motions plotted in terms of peak ground
acceleration (PGA) for Test2, Test 4D, and Test 6. As shown in Fig. 2.59, the ratio
decreases with increasing motion intensity. The ratio ranges from 1.04 in Test 2 to 0.74
in Test 6. In the higher intensity test, much damage was noticed at the cap beam and
more relative movement between the top of the column and the DT7 location is probable.
In addition, it is important to note that the quality of images decreases with increasing
earthquake intensity and incompatibilities in recording frequency between the cameras
and the earthquake motions occurred more frequently.
1.2
1
Ratio
0.8
0.6
0.4
Ratio
0.2
0
0
0.2
0.4
0.6
0.8
1.2
1.4
PGA (g)
Fig. 2.59 Maximum lateral movement ratios between photogrammetry and LVDT results
.
112
Hinging Regions
The displacement calculations are repeated at each of the grid points along the
hinging surface. Three vertical lines, as defined by the grids shown in Fig. 2.17, are
monitored through each test. On the middle line, the selected Points are 3, 8, 13, and 18
at the bottom grid and 38, 45, 52, and 59 at the top grid system. For the left line, Points
2, 7, 12, and 17 are chosen at the bottom and Points 37, 44, 51 and 58 are selected at the
top. For the right line, Points 4, 9, 14, and 19 and 39, 46, 53, and 60 are calculated at the
bottom and top, respectively.
Plots illustrating the calculated deformed shapes in the hinging region for Test 2
are shown in Figs. 2.60, 2.61, and 2.62. The overall deformed shape of the column for
Test 2 at the maximum column drift is shown in Fig. 2.60. The deformations in the
hinging regions are compared in Figs. 2.61 and 2.62 with the associated column images
at the time of maximum column drift. The deformations (Figs. 2.61-2.62) correlate well
with the associated images. For Test 2, little nonlinear deformation is noted.
113
80
Right Line
70
60
50
40
30
20
10
0
-6 -4 -2 0 2 4 6
Drift (in.)
Fig. 2.60 Overall deformed shape along the column height at maximum column drift
.
114
Middle Line
Left Line
Right Line
71
69
67
65
63
61
59
-6
-4
-2
Drift (in.)
Fig. 2.61 Top grid deformed shape with picture comparison at maximum column drift
Bottom Grid Deformed Shape (Test2)
16
14
12
10
4
Middle Line
Left Line
Right Line
0
-6
-4
-2
Drift (in.)
Fig. 2.62 Bottom grid deformed shape with picture comparison at maximum column drift
.
115
The same calculations are repeated for Test 4D (Table 2.1). Figure 2.63 shows
the overall deformed shape of the column at the maximum column drift. Reverse
curvature is noted in the top and bottom hinging regions in Fig. 2.63. The deformations
in the hinging regions are compared in Figs. 2.64 and 2.65 with the associated column
images at the time of maximum column drift. For Test 4D, nonlinear deformation is
noted very clearly in the top hinging region.
116
Middle Line
Left Line
Right Line
70
60
50
40
30
20
10
0
-6 -4 -2 0
2 4
Drift (in.)
Fig. 2.63 The overall deformed shape along the column height at maximum column drift
.
117
Middle Line
Left Line
Right Line
71
69
67
65
63
61
59
-6
-4
-2
Drift (in.)
Fig. 2.64 Top grid deformed shape with picture comparison at maximum column drift
.
Middle Line
Left Line
18
Right Line
16
14
12
10
8
6
4
2
0
-6
-4
-2
Drift (in.)
Fig. 2.65 Bottom grid deformed shape with picture comparison at maximum column drift
.
118
The overall deformed shape of the column at the maximum column drift is shown
for Test 6 in Fig. 2.66. Reverse curvature is profoundly noted for Test 6, similar to the
deformed shape for Test 4D. The deformations in the hinging regions are compared in
Figs. 2.67 and 2.68 with the associated column images at the time of maximum column
drift. The deformations represent the images very well.
119
80
70
60
50
40
30
20
10
0
-6 -4 -2 0 2 4 6
Drift (in.)
Fig. 2.66 The overall deformed shape along the column height at maximum column drift
.
120
Middle Line
Left Line
Right Line
73
71
69
67
65
63
61
59
-6
-4
-2
0
Drift (in.)
Fig. 2.67 Top grid deformed shape with picture comparison at maximum column drift
Middle Line
Left Line
18
Right Line
16
14
12
10
8
6
4
2
0
-6
-4
-2
0
2
Drift (in.)
Fig. 2.68 Bottom grid deformed shape with picture comparison at maximum column drift
121
2.3.4.3 Rotations
As described previously, several methods were evaluated to define the unique
points to calculate the displacements and rotation results from photogrammetry data.
First, using only the even-numbered lines from each frame was used to capture a more
consistent image (without double lines) for analysis. Then, a definition of vertical and
horizontal grid lines was determined. In the first definition, the mid-point of each 0.25in. (6 mm)-thick vertical and horizontal grid line was sought and a middle line was
defined as the connection of these points. In the second definition, the edge lines were
analyzed to obtain the four corner points at each intersecting thick grid line, and also the
average middle point between the corner points. In addition, the length used to define the
edge lines was examined in detail. After all these trials, using the even-numbered lines
from each image, with the average middle of the corner points, with 0.5 in. (13 mm) 1.5
in. (38 mm) in length provided the most consistent results. Using this approach defined
by the edge-lines of the thick grid lines to calculate the rotations of horizontal line at a
given height of the column, there is a large amount of pixelization noted (Fig. 2.51) and
it is difficult to follow the periodicity of the earthquake shakings. By contrast, rotations
of vertical lines provided smoother traces than the rotations of horizontal lines in the
photogrammetry data (Fig. 2.52). It is important to note that the rotations of vertical
lines are average rotations over a given column height defined by the monitored Points.
This section will describe the best final rotation results.
The rotations of vertical lines calculated using Points 3 and 8 and Points 8 and 13
on the bottom grid are shown in Figs. 2.69 and 2.70 for Test 2. There is a small amount
of constant noise (pixelization) in the figures because the amplitude of motion in Test 2
is small. Figures 2.71 and 2.72 show the rotations of vertical lines at the top grid level
122
(Points 38 and 45 and Points 45 and 52) during Test 2. (The general periodicity of
motion can be decerned among the constant noise at low amplitudes of motion.)
0.004
0.003
0.002
Rotation (rad)
0.001
0
0
10
15
20
25
30
35
-0.001
-0.002
-0.003
-0.004
Time (sec.)
Fig. 2.69 Rotation of vertical line calculated using Point 3 and 8 for Test 2
0.01
0.008
0.006
Rotation (rad)
0.004
0.002
0
0
10
15
20
25
30
-0.002
-0.004
-0.006
-0.008
Time (sec.)
Fig. 2.70 Rotation of vertical line calculated using Point 8 and 13 for Test 2
123
35
0.008
0.006
0.004
Rotation (rad)
0.002
0
0
10
15
20
25
30
35
-0.002
-0.004
-0.006
-0.008
-0.01
Time (sec.)
Fig. 2.71 - Rotation of vertical line calculated using Point 38 and 45 for Test 2
0.01
0.008
0.006
Rotation (rad)
0.004
0.002
0
0
10
15
20
25
30
-0.002
-0.004
-0.006
-0.008
-0.01
Time (sec.)
Fig. 2.72 - Rotation of vertical line calculated using Point 45 and 52 for Test 2
124
35
The rotations of vertical line calculated from Points 3 and 8 and Point 8 and 13
for Test 4D are shown in Figures 2.73 and 2.74. Figures 2.75 and 2.76 show the
rotations of vertical line at the top grid level for Test 4D.
Figures 2.77 and 2.78 show the results for the bottom hinging region for Test 6,
and Figures 2.79 and 2.80 show for the top hinging region. Both figures provide similar
peak values and periodicity.
0.02
Rotation (rad)
0.015
0.01
0.005
0
0
10
15
20
25
30
-0.005
-0.01
Time (sec.)
Fig. 2.73 - Rotation of vertical line calculated using Point 3 and 8 for Test 4D
125
35
0.03
0.025
0.02
Rotation (rad)
0.015
0.01
0.005
0
-0.005
10
15
20
25
30
35
-0.01
-0.015
-0.02
Time (sec.)
Fig. 2.74 Rotation of vertical line calculated using Point 8 and 13 for Test 4D
0.03
0.025
0.02
0.015
Rotation (rad)
0.01
0.005
0
0
10
15
20
25
30
-0.005
-0.01
-0.015
-0.02
-0.025
-0.03
Time (sec.)
Fig. 2.75 Rotation of vertical line calculated using Point 38 and 45 for Test4D
126
35
0.03
0.025
0.02
0.015
Rotation (rad)
0.01
0.005
0
0
10
15
20
25
30
35
-0.005
-0.01
-0.015
-0.02
-0.025
-0.03
Time (sec.)
Fig. 2.76 Rotation of vertical line calculated using Point 45 and 52 for Test4D
0.05
0.04
Rotation (rad)
0.03
0.02
0.01
0
0
10
15
20
25
30
-0.01
-0.02
-0.03
Time (sec.)
Fig. 2.77 Rotation of vertical line calculated using Point 3 and 8 for Test 6
127
35
0.06
0.05
0.04
Rotation (rad)
0.03
0.02
0.01
0
0
10
15
20
25
30
35
-0.01
-0.02
-0.03
Time (sec.)
Fig. 2.78 Rotation of vertical line calculated using Point 8 and 13 for Test 6
0.08
0.07
0.06
Rotation (rad)
0.05
0.04
0.03
0.02
0.01
0
-0.01 0
10
15
20
25
30
-0.02
-0.03
Time (sec.)
Fig. 2.79 Rotation of vertical line calculated using Point 38 and 45 for Test 6
128
35
0.05
0.04
Rotation (rad)
0.03
0.02
0.01
0
0
10
15
20
25
30
35
-0.01
-0.02
-0.03
Time (sec.)
Fig. 2.80 Rotation of vertical line calculated using Point 45 and 52 for Test 6
rotations using the photogrammetry data closely match the maximum values calculated
from cross-sectional rotations (Fig. 2.55).
5
4.5
5 in.-Bottom
max-PH /max-LVDT
12 in.-Bottom
5 in.-Top
3.5
12 in. Top
3
2.5
2
1.5
1
0.5
0
0
0.5
1.5
PGA (g)
Fig. 2.81 Maximum rotation ratios between photogrammetry compared LVDT results
.
2.4
CONCLUSION
For each deformation calculation, it is necessary to have a consistent and
predictable definition of points on the column surface. The quality of images captured
during the test and the similar pixel intensities across the grid line thickness presented
challenges for calculating consistent displacement and rotation results. The following
conclusions are made regarding the analysis of the deformations in the hinging region
during response to strong ground shaking:
The best image quality is obtained by using the even-numbered lines from
each frame in time.
130
The edges of the 0.25-in. (6 mm) thick grid lines are best defined using a
constant Robert threshold value.
The edge of the vertical and horizontal grid lines should be much sharper,
with distinct change in intensity from the black to white regions in the
grid.
131
Cameras that operate using progressive scan rather than interlaced images
when recording the events should be used.
132
1147 Learned Hall, Department of Civil, Env, and Arch. Engineering, University of Kansas, Kansas,
USA
2150 Learned Hall, Department of Civil, Env, and Arch. Engineering, University of Kansas, Kansas,
USA
c
Corresponding author Tel: +1 785 864 3760 Fax: +1 785 864 5631 E-mail addresses: zeynep@ku.edu (Z.
F. Alemdar).
ABSTRACT:
The determination of the location of nonlinear response in structural systems is
an important step to predict the performance of the system under different loading
conditions. In bridge columns, these nonlinear deformations generally occur over a finite
hinge length. A model of hinging behavior in reinforced concrete bridge columns will
help guide proportioning, detailing and drift estimates for performance-based design.
Data was collected during the NEESR investigation of the seismic performance of fourspan large-scale bridge systems at the University of Nevada Reno that details
deformations in column hinging regions during response to strong shaking events. A
photogrammetry method was applied using a reference grid on the top and bottom
column surfaces to accurately record and analyze deformations in the plastic hinging
regions. The method of application in this study holds several advantages over traditional
sensors including that it is a remote visualization technique, inexpensive and simple to
133
analyze the results. The surface deformations and rotations of a reinforced concrete
bridge column under dynamic loading has been examined and compared with the results
obtained from traditional instruments.
1. Introduction
134
2. Experimental program
University of Nevada Reno (UNR) laboratory on February 12-15, 2007 was significant
because it was a large-scale bridge system with columns in double curvature response. In
addition, massive earthquake motions were applied in the longitudinal and transverse
directions of the bridge.
Traditional instrumentation that have been widely used to collect deformation
data from static and dynamic tests include LVDTs (linear variable differential
transformers), displacement transducers, accelerometers, and strain gages. These
instruments are attached to the concrete surface and require electrical connection to a
data acquisition system. As damage occurs to concrete elements, the cover concrete can
spall, gages can fail, and data may be lost. A newer method used to collect deformation
data is by photogrammetry with an advanced non-contact measurement system such as
Krypton system. These systems consist of a suite of cameras, a set of infrared LEDs
(light-emitting diodes), and control software. For data acquisition during testing, the
Krypton cameras are mounted on a reference frame away from the specimen and LEDs
are attached to the specimen surface. The three-dimensional motions of infrared LEDs
(light-emitting diode) are tracked during testing. Several studies have been completed
using the Krypton LED-based 3D displacement measuring system [18-21]. Reducing and
organizing the large amount of data that is collected from this system, hovewer, is
135
difficult. Other limitations of using such advanced systems include the possibility of lost
LEDs during testing due to specimen damage, and the prohibitive cost of the system for
many structural engineering laboratories.
The four-span large scale reinforced concrete bridge test provided a platform to
compare traditional instrumentation measurements with those obtained using a simple
photogrammetry method. The photogrammetry method required only the application of a
reference grid on the concrete surface and a system of cameras to record the motion of
the bridge system. The photogrammetry data showed that the accuracy of the
measurements was approximately 0.04 in. (1 mm). The application of the
photogrammetry method held several advantages over traditional sensors. The cameras
were more economical than those used in other advanced non-contact measuring systems
and could be used in any laboratory. The photographic data did not rely on any physical
connection to the specimen. In addition, any concrete marking that were lost from
excessive deformations could be reconstructed to a certain extent based on surrounding
elements in the frames.
The test specimen consisted of the four-span reinforced concrete bridge with end
abutments as shown in Fig. 1. The bridge was quarter-scale with two 29-ft (8.84 m)
interior spans, and two 24.5-ft (7.47 m) exterior spans for a total length of approximately
110 ft (33.5 m). The clear heights of the bents were 5, 6 and 7 ft (1.52, 1.83, and 2.13 m),
with the tallest bent in the middle (Bent 2). The superstructure consisted of a solid slab
that was post-tensioned in both the longitudinal and transverse directions. The bridge
136
also included abutment seats at both ends that were driven in the longitudinal direction
by dynamic hydraulic actuators. The depth of the cap beams was 15 in. (0.38 m). The
total length of the cap beams (perpendicular to the longitudinal axis of the bridge) was 98
in. (2.49 m). The concrete compressive strength used in the bridge (excluding the slab)
was 6.7 ksi (46.2 MPa) and reinforcement was ASTM A706 Grade 60. The bents were
numbered Bent 1, Bent 2, and Bent 3 starting from left (South) to right (North).
N
SHAKE TABLE
Abutment 1
SHAKE TABLE
Bent 1
Bent 2
SHAKE TABLE
Bent 3
Abutment 2
Photogrammetry data was collected at the top and bottom hinging regions of the
east column of Bent 3 during the tests. Two grid systems were applied to the surface of
the column as shown in Fig. 2. The bottom grid consisted of vertical and horizontal lines
spaced irregularly around the face of the circular column. Deformation data was
collected at various points of the intersecting lines. The top grid system used a
combination of lines and squares, having the same spacing as the lines in the bottom
grid. This top grid pattern was applied to simulate targets on the column surface.
137
To apply the grid systems, the column was first spray-painted black for a solid
dark background. The lines and squares were lightly traced with a pencil, and then a
single stencil was used to fill in the lines and squares using white paint.
(a)
(
b)
Fig. 2 Close view of (a) bottom and (b) top grid systems
138
DT7
3ETR3
57
56 58
50
49 51
59
52
62
6061
55
5354
43
42 44
45
48
4647
3ETR5
3ETR6
36
35 37
29
28 30
212223
41
38 3940
34
31 3233
24 252627
1617 18
19 20
1112
13
1415
6 7
9 10
1 2
4 5
3EBR5
3EBR7
3ETR4
3EBR6
3EBR8
Fig. 3 Grid system and LVDT locations on column in the Bent 3 east column
An aluminum tower was built to hold the cameras that would record the column
deformations caused by the earthquake events (Fig. 4). The tower had two levels to allow
recordings at the top and bottom plastic hinging regions. A system of two cameras
(DXB-9212EF model Starlight 600 TVL super high resolution) with black-white
recording properties was mounted on each level [at 94.5 in. (2400 mm) and 159 in. (4039
mm) height]. Video-lens zooms had 60-300 mm F4-5.6 capturing capability.
139
The motions that were used for the shaking table tests were based on the 1994
Northridge earthquake as measured at the ground station at the Century City Country
Club North. The Century City station is owned by the California Strong Motion
Instrumentation Program, and is located at 34.063 latitude, -118.418 longitude. Both the
90 degree and 360 degree lateral components were used in the tests. The same motion
was used for each test, but the peak ground acceleration (PGA) of the motion was
amplified as additional tests were completed. Over the course of the tests, the PGA of the
original motion was scaled from 0.075g to 1.0g in transverse direction (90 degree
component) and from 0.09g to 1.2g in longitudinal direction (360 degree component) as
140
shown in Table 1. The earthquake motion was applied for uniaxial and biaxial tests of
the structural system. The motion was applied 13 times with six different scaled
intensities so that the progression of damage would be evaluated from pre-yield to failure
during the shaking tests.
Table 1 Earthquake motion levels applied during the tests
Test No.
Test Date
1A
Motion
Level
Test Type
Long.
Longitudinal
0.09
Longitudinal
0.09
1C
Longitudinal
0.09
1D
Biaxial
0.075
0.09
Biaxial
0.15
0.18
Biaxial
Longitudinal
0.6
4B
Longitudinal
0.6
4C
Longitudinal
0.6
4D
Biaxial
0.5
0.6
Biaxial
0.75
0.9
Biaxial
1.2
Biaxial
1.2
1B
4A
5
6
6
2/12/2007
2/13/2007
2/14/2007
2/15/2007
141
analysis represented the rotations of horizontal lines based on the movement of the
column, which included flexural and rigid body components. Rotations calculated using
the LVDT measurements from the test included only flexural contributions. Another
difference was that the LVDT instrumentation was set to record movement in the
transverse and longitudinal planes, whereas the photogrammetry measurements were
recorded at a horizontal angle between the orthogonal directions, as selected with the
setup of the aluminum tower and the cameras (Fig. 4). These differences are discussed
and accounted for in the analysis and results.
A measure of the best definition of unique points and deformations was made
using a FDE (frequency domain error) index as developed by Dragovich and Lepage
[22]. The FDE index is used to compare time-domain response waveforms such as those
described in this study. The FDE method uses the Fourier spectra of the response signals,
which are non-periodic and comprised of multiple frequencies, to calculate the index.
Fast Fourier Transform was used to determine the composition of frequencies of the
photogrammetry responses. The starting and ending frequencies were selected to be
between 0.5 Hz and the Nyquist frequency, which is fn =1/(2t) with t as the time
interval of the series. The FDE index accounts for both the amplitude and phase
differences between the response signals and gives a number between 0 and 1 as a
measure of the goodness-of-fit, where zero indicates a perfect correlation. The amplitude
error is a measure of the difference between the values of the response signals, and the
phase angle is the angle that the resultant vector of the signals makes with respect to the
real axis in the real-complex plane. From the study by Dragovich and Lepage, an FDE
value equal to 0.75 represented a poor correlation, and an FDE value of 0.25 represented
a very good correlation. These correlations, however, were based on analytical responses
142
Interactive Data Language (IDL) was used to analyze the video images that were
recorded during the test. Each video was first divided into over one thousand images for
individual analysis. The displacement of a unique point in an image was defined as the
relative horizontal movement of that point between successive images. At times,
however, it was difficult to keep track of a unique point because of the quality of the
images.
Each image (640 x 480) was written as 480 lines that were 640 pixels wide, and
represented 1/30th of a second during the test. The images were interlaced: the oddnumbered lines are written first, followed by the even-numbered lines so that the image
was really two images offset by 1/60th of a second. The first partial image (half of the
total image) was comprised of the odd lines, and the second partial image was
comprised of the even lines. Due to shaking of the strong floor during the tests and the
incompatibility in recording frequency between the cameras and the earthquake motions,
several images were degraded in quality (corrupted). As a result, some of the horizontal
and vertical lines appeared as double lines, and some images were simply blurry. In
future tests, cameras that operate with progressive scan rather than interlaced images
should be employed to avoid these problems.
The even-numbered and odd-numbered lines of the images were considered
separately to improve the calculations of displacements and rotations with the corrupted
images. When the image was divided into separate images of even and odd lines, the
143
double image effect that was produced by a shift between the two halves of the image
being written was eliminated. The vertical displacement results of Point 7 (located in Fig.
3) from the even and odd line analyses were reduced from the Test 4D motion (Table 1).
A comparison of the calculated response obtained from the photogrammetry
analysis and the measured response from traditional instruments was completed. The
calculated photogrammetric response was sensitive to how unique points on the column
surface were defined. The quality of the images during the shaking events affected the
final representation of the column deformations. The vertical displacement results of
Point 7 are compared with the LVDT 3EBR7 readings in Figs. 4 and 5 for even and odd
line analyses, respectively. FDE indexes were calculated for the vertical displacements of
Point 7 in the even and odd analysis as compared to LVDT 3EBR7 results. Table 2
shows that the amplitude error is very good at only 0.16 for the even line analysis,
whereas it is 0.19 for the odd line analysis. The phase errors (periodicity errors) are very
similar at around 0.50. Overall, the total FDE index is slightly smaller for the even line
analysis than the odd lines, and is indicative of which set of lines (even/odd) captured
first.
Table 2 FDE index results for Point 7
FDE Results
Comparisons
Fig. 7-Even line
analysis vs. LVDT
Fig. 8-Odd line
analysis vs. LVDT
Amplitude
Phase
Total
0.16
0.51
0.67
0.19
0.49
0.68
0.12
0.58
0.70
0.10
0.45
0.55
144
Fig. 5 Comparison of Point 7 vertical displacement for even lines with LVDT 3EBR7
data
Fig. 6 Comparison of Point 7 vertical displacement for odd lines with LVDT 3EBR7 data
145
The definition of each unique point on the grid surface was a critical step to
obtain the displacements of the points and rotations of sections of the column. The
intersection of the thick vertical and horizontal lines (0.25 in. thick) was used to define
the points, however, there were several ways to consider creating a line within that
thickness. One option was to find the mid-point of each 0.25-in. thickness and define a
middle line with the connection of those points (Fig. 6). The pixel that had the
maximum intensity value in the intersections of the horizontal and vertical lines was
defined as the mid-point within the thickness.
The edges of the thick grid lines were considered as a second option to define two
new lines (edge-lines) for each horizontal and vertical grid line. Using this option, there
were four intersections of the edge lines to consider as possible unique locations of that
Point, as seen in Fig. 7. In this method, the edge lines may be determined by points of
a particular light intensity value, a Robert Threshold, or as a considerable difference in
relative intensity between the black to white paint transition. The length of the line also
had an effect on the results. Longer [1.5 in. (38 mm) to 3.0 in. (76 mm)] and shorter [0.5
in. (13 mm) to 1.5 in. (38 mm)] edge-line lengths were investigated to determine the
influence on the definition of a unique Point.
The best representation of a unique Point was obtained from the average of four
corner point coordinates. First, the four corner points at the intersection of each
horizontal and vertical grid line were defined using the edge-lines of the 0.25-in.-thick
grid lines. These edge lines were defined using a constant Robert threshold value of 50
with a short edge line length that ranged from 0.5-1.5 in. The coordinates of the four
corner points were then averaged to define a middle point.
146
Middle Line
Edge Lines
Edge Point
Middle Point
13
12
c
147
points within each image was accounted for in the analysis. The rotation of a vertical line
on the column was approximated using the relative transverse displacements of two
points as reduced from photogrammetry data that were aligned vertically on the grid
surface divided by the length between those lines. This rotation represents the average
cross-sectional rotation between those two points if shear deformations can be neglected.
Figure 8 shows the rotation of a vertical line that was calculated using the lateral
displacements of Points 3 and 8 on the bottom grid as were tracked during the Test 4D
motion. Figure 9 shows the rotation of a horizontal line located at the top of the region
encompassed by the vertical line (calculated using the vertical displacements of Points 7
and 9).
The comparison between the rotations of the vertical line shown in Fig. 8 and the
horizontal line shown in Fig. 9 is shown in Fig. 10. Although the top and bottom hinging
regions had numerous cracks and some spalling of the concrete on the column surface, to
the extent that a few lateral reinforcing bars and yielding of the longitudinal bars were
revealed, the photogrammetry method was able to capture the rotations well. As an
average measurement, the rotation of the vertical line provides a smoother trace than the
rotation of the horizontal line that was calculated with the photogrammetry data. The
correlation between the rotation of vertical and horizontal line was examined using the
FDE index (Table 2), and the amplitude error was found to be 0.12, which represents a
very good correlation. The phase error, however, was 0.58 and the total FDE index was
0.70, which overall was a fair correlation.
148
0.03
0.025
0.02
Rotation (rad)
0.015
0.01
0.005
0
-0.005
-0.01
-0.015
-0.02
0
10
15
20
Time (sec)
25
30
35
40
35
40
0.03
0.025
0.02
Rotation (rad)
0.015
0.01
0.005
0
-0.005
-0.01
-0.015
-0.02
0
10
15
20
25
30
Time (sec)
149
0.03
Rotation of a Horizontal Line
0.025
0.02
Rotation (rad)
0.015
0.01
0.005
0
-0.005
Rotation (rad)
-0.01
-0.015
-0.02
0
10
15
20
Time (sec)
0.03
0.02
0.01
0
-0.01
-0.02
25
30
8
10
35
T ime (sec)
150
system (Fig. 3), which was the closest point to the top of the column, was selected to
track the transverse displacement along the direction of the focal plane of the cameras on
the left side of the aluminum tower. The projection of the camera view from the tower
had an angle of 15.25 with the longitudinal direction of the bridge (Fig. 4). The
transverse displacements of grid Point 59 captured during Test 4D was compared with
the combined components of the displacement transducers mounted on the bridge deck
and the north abutment as shown in Fig. 11 and described next. First, the measurements
recorded by DT7 (transverse displacement of Bent 3 at mid-height of the bridge deck)
were used to obtain the transverse component along the focal plane of the cameras. Then,
DS5 and DS1 measurements were combined together to find the longitudinal component
of the displacement (DS5 recorded the gap size between the north abutment and the
bridge deck while DS1 measured the movements of the north abutment in the
longitudinal direction of the bridge). Although there was approximately 23.3 in. (592
mm) between the transverse beam and Point 59, the column deformed in double
curvature and the lateral displacements at these two heights were assumed to be similar.
As seen in Fig. 11, the overall periodicity of the two measurements correlates very well,
with differences in magnitude primarily attributed to small flexibilities of the transverse
beam creating differential movement between the top beam and Point 59 in the column
hinging region.
The FDE index analysis was used to compare the results of transverse
displacements at Point 59 and the movement recorded by the displacement transducers.
The amplitude error was found to be 0.10, which represented a very good correlation
between the photogrammetry and displacement transducers results (Table 2). This is
significant as the maximum displacement of bridge columns is an important parameter
151
used to qualify expected damage during earthquake events. The phase angle error was
calculated to be 0.45, which was a good correlation result.
The deformed shapes of the hinging regions were defined over a continuum by
considering the movements of a suite of Points in the regions [23]. The displacement
calculations were repeated at each of the grid points along the hinging surface. Three
vertical lines, as defined by the grids shown in Fig. 3, were monitored in Test 4D. On the
middle line, the selected Points were 3, 8, 13, and 18 for the bottom grid and 38, 45, 52,
and 59 for the top grid system. For the left line, Points 2, 7, 12, and 17 were chosen at
the bottom and Points 37, 44, 51 and 58 were selected at the top. For the right line,
displacements of Top Points 4, 9, 14, and 19 and Bottom Points 39, 46, 53, and 60 were
monitored. The deformations in the top and bottom hinging regions during Test 4D, as
represented by dotted lines connecting the displaced points, are compared in Figs. 12
and 13 with the associated column images at the time of maximum column drift. Reverse
curvature is correctly represented in the top and bottom hinging regions. Slight nonlinear
deformation is also visible in the top hinging region. The deformations captured by the
photogrammetry data represent the images very well as seen in Figs. 12 and 13.
152
80
Photogrammetry
DT7 & (DS5+DS1) combined
components
60
Displacement (mm)
40
20
0
0
10
15
20
25
30
35
40
-20
-40
Time (sec.)
1850
1800
1750
1700
1650
1600
1550
1500
-50
50
100
150
Fig. 13 Top grid deformed shape with picture comparison at maximum column drift
during Test 4D
153
450
400
350
300
250
200
150
100
50
-100
-50
50
100
150
Fig. 14 Bottom grid deformed shape with picture comparison at maximum column drift
during Test 4D
4. Conclusions
154
The FDE indexes obtained for the vertical displacements of Point 7 in the
even and odd analysis gave very good correlations between the calculated
amplitude using photogrammetry data and the measured responses from the
LVDT 3EBR7 recordings. The phase error results had fair FDE values
(approximately 0.50), and fair overall values (0.67 and 0.68).
The deformed shape of the top and bottom hinging regions was constructed
using the displacement of the points on the column surface. The deformed
shapes show a good match with the images at the maximum drift of Test 4D
by correctly representing double curvature of the column and the general
deformed shape of the hinging region.
155
Acknowledgements
This project gratefully acknowledges the support of NSF grant #0532084, Joy
Pauschke Program Director. The assistance and support of Prof. Saiid Saiidi, Graduate
Research Assistant Roby Nelson, the entire staff of the research facilities at the
University of Nevada Reno, and Graduate Student Nick Hunt at University of Kansas are
also gratefully acknowledged. Prof. Olafsen was supported in part by a Big XII Faculty
Fellowship. Additional support from KU Transportation Research Institute is gratefully
acknowledged.
References
156
157
[16] Dodd L.L. The dynamic behavior of reinforced concrete bridge piers subjected
to New Zealand seismicity. Ph.D. thesis, the University of Canterbury Christchurch
New Zealand; 1992.
[17] Hachem M. M., Mahin S.A., Moehle J.P. Performance of circular reinforced
concrete bridge columns under bidirectional earthquake loading. Report No. PEER
2003/06, Pac. Earthq. Eng. Research Center, University of California at Berkeley,
Feb. 2003, 490 pp.
[18] Brueggen B., Waugh J., Aaleti S., Johnson B., French C., Sritharan S. et
al. Tests of structural walls to determine deformation contributions of interest for
performance-based design. Proc. of 2007 ASCE Structures Congress, Long Beach,
CA, 2007.
[19] Matamoros A. B., Matchulat L., Woods C. Axial load failure of shear critical
columns subjected to high levels of axial load. Proc. 14th World Conference on
Earthquake Engineering. 2008.
[20] Matchulat Lisa. Mitigation of collapse risk in vulnerable concrete buildings,
M.E., University of Kansas; 2009.
[21] Leon R.T., Kim D.K., Hajjar, J.F. Limit state response of composite columns
and beam-columns. Part 1: Formulation of design provisions for the 2005 AISC
Specification, AISC Engineering Journal, Vol. 44, No. 4, pp. 341-358. 2007.
[22] Dragovich JJ., Lepage A. FDE index for goodness-of-fit between measured and
calculated response signals. Earthquake Engineering and Structural Dynamics
2009;38:1751-1758.
[23] Firat Alemdar Z. Evaluation of plastic hinge regions in reinforced concrete
bridge systems. Ph.D. dissertation, Lawrence (KS): the University of Kansas; 2010.
158
a,*
Zeynep Firat Alemdar , Adolfo Matamorosb and JoAnn Browningc
1147 Learned Hall, Department of Civil, Env, and Arch. Engineering, University of Kansas, Kansas,
USA
2150 Learned Hall, Department of Civil, Env, and Arch. Engineering, University of Kansas, Kansas,
USA
2150 Learned Hall, Department of Civil, Env, and Arch. Engineering, University of Kansas, Kansas,
USA
Corresponding author Tel: +1 785 864 3760 Fax: +1 785 864 5631 E-mail addresses: zeynep@ku.edu (Z.
F. Alemdar).
1. Introduction
160
the column. The displacements were reduced using the pictures recorded by the cameras.
The bridge column was analyzed using Finite element (FE) analysis with the ABAQUS
[12] software. Analyses of the bridge column under monotonic and earthquake-induced
deformations were completed, and the load deflection curve under monotonic loading
was compared with a similar curve produced using the Modified Compression Field
Theory. The FE model results under earthquake-induced deflections were compared with
the photogrammetry measurements.
161
The top and bottom hinging regions of the east column of Bent 3 were the main
targets to collect photogrammetric data in the large-scale four-span reinforced concrete
bridge test. The column that was selected for application of the photogrammetry method
was 72 in. (1829 mm) long and 12 in. (305 mm) in diameter, with a 0.5 in. (13 mm)
concrete cover. The column consisted of 16 No.3 (16-10M) longitudinal reinforcement
equally spaced along the radial direction. The cap beam is 15 in. (381 mm) by 15 in. with
a 23- in. (584 m) width. The total length of the cap beams along the transverse direction
of the bridge was 98 in. (2490 m).
Two different grid systems were applied to track the displacements of the
columns (Fig. 2). To construct these grid systems, the bridge column was first spraypainted black to obtain a solid dark background and the lines and squares were lightly
traced with a pencil, and then they were filled in using a single stencil and white paint.
The bottom grid had vertical and horizontal lines spaced irregularly around the face of
the circular column while the top grid used a different configuration that was composed
of lines and squares so as to be similar to the spacing of the lines in the bottom grid. The
intersections of the thick vertical and horizontal lines in both regions were numbered and
used to obtain the best definition of Points along the column surface after the analysis of
the photogrammetry data as shown in Fig. 3 [13].
The biaxial and uniaxial scaled motions measured at the Century City Country
Club during the 1994 Northridge, California earthquake were applied during the bridge
tests. This earthquake motion was used 13 times with increasing amplitude in order to
monitor the progression of damage from pre-yield to failure. A white noise excitation
was also applied before each ground motion intensity change to test for system
162
A three dimensional finite element model of the east column of Bent 3 was
defined with the cap beam on the top of the column and the footing system under the
circular column using ABAQUS Finite Element software. A concrete damaged plasticity
model was utilized in the FE analysis to define the uniaxial compressive and tensile
concrete material properties of the bridge column, while elastic concrete material
properties were assumed for the cap beam and footing of the system. The properties of
163
the concrete in the core and cover were defined using a plasticity based material model.
The stress-strain curve in compression was defined using the Mander unified stress-strain
model under monotonic loading at slow strain rates [14] as shown in Fig. 8. The
maximum strain was selected to be 0.1, similar to the research done by Scott et al.
(1982), Spacone et al. (1996), and Yassin (1994) who used the maximum concrete strain
in compression as 0.06. The response of concrete subjected to the moderately rapid load
rates, as associated with earthquake ground motions, was also considered. The dynamic
magnification factors that were used in the analysis followed the recommendations of
Mander (1984), who used regression analyses of the experimental results of Watstein
(1953) on plain concrete specimens of different strengths. The relationships for the
dynamic strength of concrete (Df) are as follows.
(fco)dyn = Df fco
(1)
where:
(fco)dyn = the compressive strength of concrete under rapid loading
fco = the quasi-static compressive strength of concrete in MPa (1 Mpa=145 psi).
1/ 6
c
1+
0.035( f co' ) 2
Df =
1/ 6
0.00001
1+
' 2
0.035( f co )
(2)
164
(3)
1/ 6
c
1+
' 3
0.035( f co )
DE =
1/ 6
0.00001
1+
' 3
0.035( f co )
(4)
165
A uniaxial bilinear steel model with isotropic kinematic hardening properties was
used to simulate the behavior of the longitudinal steel. Isotropic parameters were
implemented using the classical stress-strain relationship for ductile steel according to
ASTM A706 Grade 60 Steel. In order to adjust the steel strength for loading in the range
of strain rates developed under earthquake shakings, dynamic magnification factors were
calculated. The data for A615 steel [21] showed that the dynamic yield strength was
approximately 10 percent larger than the static yield strength. According to
investigations done by Manjoine (1944), the maximum increase in ultimate strength of
mild structural steel was approximately 4%. Several studies [21, 25, and 26] indicated
that increased strain rate had relatively little effect on the elastic modulus of steel. The
dynamic factor of 10 percent for the yield strength was used in the material model and
compared as shown in Fig. 10.
Kinematic hardening properties were defined including cyclic strain softening as
defined by Ma et al. (1976). A reinforcing steel element was modeled with the cyclic
material properties that had been proposed by Ma et al. (1976) and subjected to reversed
cyclic loading. The stress-strain curve of an element along the steel member was
obtained from the FE analysis and compared with the experimental data conducted by
Ma et al. (1976) as shown in Fig. 11. The FE model represented the results reported by
Ma et al. (1976) very well.
Mesh sensitivity analysis was studied under monotonic loading for the FE model
of the bridge column. The axial load on the column consisted of an imposed weight of
47.2 kips applied as a uniform pressure at the surface of the cap beam in the FE model.
The two-directional shaking table displacement measurements that were recorded by the
research team at UNR were applied at the bottom surface of the footing as transverse and
166
longitudinal direction displacement histories. The circular column, cap beam and the
footing were modeled using 3D continuum 8-node brick elements, whereas longitudinal
reinforcement and transverse steel hoops were defined using 1D 2-node linear truss
elements. The longitudinal steel had a cross-sectional area of 0.11 in.2 (71 mm2) and
transverse hoops had 0.029 in.2 (18.7 mm2) cross-sectional area evenly distributed every
1.25 in. (32 mm) along the column. The brick elements had twenty integration points
distributed throughout its physical volume. A reduced integration scheme was utilized to
reduce the computation time for the analysis.
The FE mesh of the solid elements was generated considering the location of the
longitudinal reinforcement. Truss elements were linked to the edge of the solid elements
using an embedded constraint option. To determine the mesh sensitivity in the FE model,
three different models having mesh sizes of 3 in. (76 mm), 1 in. (25 mm) and 0.5 in. (13
mm) along the top and bottom hinging regions of the column and 4 in. (102 mm) outside
the regions were studied. A coarser mesh was used for the cap beam and the footing
because they were modeled with elastic material properties. The FE model was analyzed
under increasing lateral loads. Figure 12 shows that the 1 in. (25 mm) mesh model gave a
better force-displacement distribution when compared to similar curves generated using
the Modified Compression Field Theory [28] and the result of a moment-curvature
program (Curve). The analysis of the 0.5 in. (13 mm) mesh model along the hinging
regions took 2 times longer than using 1 in. (25 mm) mesh without any benefit in model
performance. The 1 in. mesh size was chosen for future analyses.
The success of the model is measured by how accurately the response of the
structure is represented. To accomplish this measurement, a quantitative procedure is
used to compare measured and calculated response values. This measure was made using
167
a FDE (frequency domain error) index as developed by Dragovich and Lepage [29]. The
FDE index is used to compare time-domain response waveforms such as earthquake
responses obtained in this study. The FDE method uses the Fourier spectra of the
response signals, which are non-periodic and comprised of multiple frequencies, to
calculate the index. Fast Fourier Transform was used to determine the composition of
frequencies of the photogrammetry responses. The starting and ending frequencies were
selected to be between 0.5 Hz and the Nyquist frequency, which is fn =1/(2t) with t as
the time interval of the series. The FDE index accounts for both the amplitude and phase
differences between the response signals and gives a number between 0 and 1 as a
measure of the goodness-of-fit, where zero indicates a perfect correlation. The amplitude
error is a measure of the difference between the values of the response signals, and the
phase angle is the angle that the resultant vector of the signals make with respect to the
real axis in the real-complex plane. From the study by Dragovich and Lepage, an FDE
value equal to 0.75 represented a poor correlation, and an FDE value of 0.25 represented
a very good correlation.
The results of the FE model of the bridge column under dynamic loading were
obtained and compared with the photogrammetric measurements. The FE model of the
Bent 3 east column was analyzed under the earthquake loading, which were selected to
have biaxial components of 0.5 g maximum ground acceleration in transverse direction
and 0.6 g maximum ground acceleration in the longitudinal direction (Test 4D). The
displacements at grid Point 59, as shown in Fig. 3, were reduced from the
photogrammetric data during Test 4D. These displacements are compared with the
combined components of the displacement transducers mounted on the bridge deck and
the north abutment, which recorded the transverse displacement of Bent 3 at the middle
168
of the bridge deck (DT7), and measured the gap size history and the displacement of the
north abutment in the longitudinal direction of the bridge (DS5 and DS1) during the
same motions. Figure 13 shows the comparison of the displacement history. The FDE
index analysis was completed to compare the results obtained from the photogrammetry
method with the measured displacement transducers. The amplitude error was found to
be 0.10, which represented a very good correlation between the photogrammetry and
displacement transducers results (Table 1). The phase angle error was calculated to be
0.45, which was a good correlation result. Although they were separated by 23 in. (584
mm), the displacements at these two points were similar because of the double curvature
along the column.
The displacement history at the top and bottom hinging regions from the FE
model were also compared with the photogrammetric data. In the bottom hinging region,
the displacement history at Point 2 as shown marked on the grid surface in Fig.3 is
plotted in Fig. 14. Point 51 (Fig. 3) was selected to compare the displacement at the top
hinging region. The periodicities of the two measurements of the motion are very similar
as seen in Figures 14 and 15. The amplitude and phase errors of the lateral displacements
(Table 1) had very good correlations, ranging from 7.3% to 9.4% in amplitude error and
from 14 % to 25% in phase error. The FE model provides very good results.
The effect of yield strength of the longitudinal reinforcement was considered
because limited information about the stress-strain relationships was available to use in
the analysis. The lateral response at the top and bottom points shown in Fig. 3 (Point 2
and 51) was insensitive to the yield strength as shown in Fig. 14 and 15. The vertical
displacement response and cross-sectional rotation were sensitive to yield strength, and
so responses were calculated using yield strength values of 64, 70, 75, and 80 ksi (441,
169
483, 517, and 552 MPa). Figures 16-19 show the comparisons of vertical displacements
at the location of LVDT 3ETR4, which was instrumented 5 in. (127 mm) below the cap
beam as shown in Fig. 3, for each yield strength. Figures 20-27 show the comparisons of
the measured cross-sectional rotations at the locations of LVDTs (LVDT 3ETR3 and
3ETR4 and LVDT 3ETR5 and 3ETR6) that were located at 5 in. (127 mm) and 12 in.
(305 mm) below the cap beam with the calculated FE analysis results. The correlations
between the vertical displacement obtained from the FE model and the LVDT
measurements were examined using the FDE index for each yield strength value as given
in Table 2. As yield strength was increased from 64 ksi to 80 ksi, the total FDE index
decreased from 31% to 18% for vertical displacement calculations. It is evident that the
yield strength of 80 ksi provides a very good comparison as shown in Fig. 28. Figures 29
and 30 show that the FDE index results for the cross-sectional rotations at 5 in. (127 mm)
and 12 in. (305 mm) are very similar to each other for all yield strength values analysis
and represent very good correlations.
The effect of the flexural stiffness of the cap beam was examined to simulate the
rotation of the cap beam during response to ground motions. The flexural stiffness of the
cap beam was calculated and modeled in the program by defining a linear-elastic
rotational spring at the center of the right face of the cap beam. The stiffness was
computed as 10^9 lb-in. (113*10^6 N-m). It was varied from 10^8 lb-in. (113*10^5 Nm) to the real value in the model to determine the effect of the flexural stiffness of the
cap beam on the response of the bridge column. Yield strengths of the reinforcement of
68 and 75 ksi (469 MPa and 517 MPa) were used in the model to find the best definition
of the cap beam stiffness. Figures 31-33 show the comparisons of the vertical
displacements from the analyses with the yield strength defined to be 68 ksi (469 MPa).
170
The stiffness values were taken to be 10^8 lb-in. (113*10^5 N-m), 1.5*10^8 lb-in.
(17*10^6 N-m), and 5*10^8 lb-in. (56.5*10^6 N-m.). The vertical displacement
comparisons are shown in Figs. 34 and 35 for the yield strength of 75 ksi (517 MPa). An
FDE index analysis was used to determine the best correlation between the analyses.
Figures 36 and 37 show that as the stiffness of the cap beam increases, the FDE
decreases. The stiffness of 5*10^8 lb-in. (56.5*10^6 N-m) gives the best comparison
when the yield strength of the steel is equal to 75 ksi (517 MPa) as shown in Table 3. In
Table 3, as the stiffness of the cap beam was increased from 1*10^8 lb-in. (113*10^5 Nm) to 5*10^8 lb-in. (56.5*10^6 N-m), the amplitude error decreased from 41% to
approximately 10%, for either fy=68 ksi (469 MPa) or fy=75 ksi (517 MPa). The phase
error with constant cap beam stiffness increased, however, from 20% to 33% with fy=68
ksi (469 MPa), and from 20% to only 29% with fy=75 ksi (517 MPa).
When reinforced concrete structures are subjected to cyclic loading, they
experience progressive damage and a reduction in stiffness. At very high deformations, it
is inevitable to expect significant damage and stiffness degradation in the bridge column,
therefore, stiffness degradation and stiffness recovery variables were used in the concrete
plastic damage model. The stiffness degradation parameters dc and dt defined in the
concrete plasticity model in ABAQUS are numerical values which depend on inelastic
compressive strain or tensile crack displacement. When the post-peak compressive stress
or tensile crack displacement increases, the material sustains non-recoverable damage,
and the stiffness of the material reduces after a load reversal occurs. In this study, the
compression stiffness degradation parameters for core and cover concrete were defined
as exponential functions dependent on the plastic component of the compressive strain.
Equations 5 and 6 were utilized for core and cover concrete respectively:
171
dc = 1 e
dc = 1 e
30 pl
(5)
160 pl
(6)
where pl is the plastic component of the compressive strain. The compression stiffness
degradation model for the core and cover concrete are shown in Figs. 38 and 39. The
compression stiffness recovery variable (wc) was taken between 0.5 and 1.0, which
implies that as cracks close during load reversal the compression stiffness will be
completely recovered.
The tension stiffness degradation parameter was defined by a linear function in
which a 50 percent reduction of the elastic stiffness was assumed to occur when the
critical crack bandwidth value was reached, and 98 percent of the elastic stiffness when
the critical crack bandwidth was doubled. Figure 40 shows the tension stiffness
degradation model. The tension stiffness recovery variable (wt) was taken to be between
0 and 1.0 in the FE analyses.
Several combinations of different compression and tension stiffness recovery
variables as given in Table 4 were defined in the model to obtain the best parameters.
The comparisons of vertical displacements are shown in Figs. 41-44. The FDE indexes
were also calculated to determine the best combinations of the parameters (Table 4). As
seen in Fig. 45, the best combination of recovery variables are wc =0.8 and wt =1.0.
These stiffness recovery variables were chosen for future analysis.
The effect of the bond-slip at the column interfaces was considered in the FE
analysis because when a column is subjected to axial load and flexure, the longitudinal
reinforcement displaces in tension and in compression as a result of the strain in the
embedded bar. The effects of the slip were modeled using linear-elastic springs at the
beam-column interface. Given the large development length afforded by the foundation
172
and cap beam this was considered to be an adequate assumption. The properties of the
springs were computed as follows. The slip displacement is calculated by integrating the
axial strains over the anchored bar length. The slip of the longitudinal reinforcement
results in additional member end rotation that is not related to flexural deformations.
Experimental results from four double curvature columns indicated that the bar slip
deformations may be as large as the flexural deformations [30]. The value of the slip
deformations did not vary much with increasing lateral displacement or damage in the
tests. Numerous experimental studies have been completed to define the anchorage
behavior of the reinforcing bars. Several analytical bond-slip models have been
developed using these experimental results. Using a bilinear strain distribution as shown
in Fig. 46, the slip in the model was ultimately determined from Eq. 7 and 8.
ld
slip = dx =
s ld
ld '
ld
s y
s ld
l 'd
( s + y )
slip = dx + dx =
+
2
2
0
ld
s > y
(7)
(8)
The development lengths over the elastic and inelastic regions of the
reinforcement (ld and ld) were calculated from the equilibrium of forces in the
reinforcement as given in Eq. 9. A uniform bond stress of ub, which was equal to 6fc,
was assumed in the elastic range (Fig. 46).
ld =
f y db
(9)
4u b
In the inelastic portion of the slip, the slope of the force-slip relationship (Fig. 47)
was assumed to be 1% of the slope of the elastic range. The total slip was calculated to
be 0.08 in. (2 mm) in the FE model having the yield strength of the longitudinal
reinforcement of 64 ksi (441 MPa).
173
The vertical displacement at the location of LVDT 3ETR4 was obtained from the
FE analysis and compared with the measured displacement at the face of the column, as
shown in Fig. 48. The vertical displacement result give higher values than those found in
Fig. 3 when the yield strength was equal to 64 ksi (441 MPa) and no bond-slip behavior
was defined in the FE model. The rotations, however, followed the measured LVDT data
very well as shown in Fig. 49 and 50. An FDE index analysis was used to determine the
effect of the added bond-slip deformation in the FE model (Fig. 51). Table 5 shows that
as the bond-slip effect was introduced in the FE model, the total FDE error increased
from 30% to 43%. At best it can be concluded that bond-slip deformations are not an
important component to the response of the bridge column.
The deformed shapes of the hinging regions were defined over a continuum by
considering the movements of selected Points in the regions [13]. Three grid lines in the
column hinging regions were selected to define the lateral displacements. These lines
were called Middle, Left, and Right lines because of their locations on the grid (Fig. 2).
Four Points on each grid line were selected to obtain the lateral displacements using the
Photogrammetric data. The lateral displacements were connected using a line to simulate
the deformed shape of the column. The lateral displacements of the same points on the
column surface shown in Fig. 3 were also determined from the FE model at the time of
the maximum drift during response to the Test 4D motion. Yield strengths of 64 and 80
ksi (441 and 552 MPa) were considered to obtain results with the FE model. Figure 52
shows the comparisons of the deformed shapes reduced from the photogrammetry data
and the FE model when the yield strength of the steel is equal to 64 ksi (441 MPa). For
the 80 ksi (552 MPa) yield strength, the deformed shapes are compared in Fig. 53. the
deformed shape obtained from the FE model with the yield strength of 80 ksi (552 MPa)
174
follows the deformed shape reduced from the photogrammetry better than the results
with the yield strength of 64 ksi (441 MPa).
4. Conclusions
The static increasing lateral load analysis results showed that using a 1 in.
(25 mm) mesh along the top and bottom hinging regions of the bridge
column provided a similar force-displacement curve as that obtained
using the Modified Compression Field Theory.
According to the analyses results using different yield strengths for the
steel, using the yield strength of 80 ksi (552 MPa) gave the best FDE
index value.
The flexural stiffness of the cap beam should be defined to be 5*10^8 lbin. (56.5*10^6 N-m).
The error calculated from the vertical displacement response including the
bond-slip effect of the longitudinal reinforcement was higher than the
error associated with the response having no bond-slip effect.
Using these modeling parameters, the FE model adequately represented the drift
response of the bridge column and plastic deformation behavior in response to strong
ground motion. The comparisons of the calculated and measured lateral displacement
175
responses of points in the hinging regions gave very good results, as calculated with the
FDE index analysis. Finally, the deformed shape of the bridge column generated using
the FE model results followed the deformed shape of the column reduced from the
Photogrammetric data.
Total
error
error
(%)
(%)
10
45
55
7.3
14
22
9.4
25
34
Amplitude
error (%)
Comparison
of Point 59
Comparison
of Point 2
Comparison
of Point 51
176
Vertical
Displacement
Rotation @ 5
in. below the
cap beam
Rotation @ 12
in. below the
cap beam
Yield
strength of
the steel
(ksi)
64
70
75
80
64
70
75
80
64
70
75
80
Amplitude
error (%)
Phase
error
(%)
Total
error
(%)
8.9
7.3
6.6
6.4
20
20
19
17
7.3
7.3
6.8
6.9
22
17
14
12
6.6
6.6
7.1
9.7
5.4
5.4
5.8
5.9
31
24
21
18
27
27
26
27
13
13
13
13
Table 3 FDE results for analyses with different stiffness of the cap beam analysis
(FE vs. LVDT vertical displacement measurements)
Flexural stiffness
Amplitude
of cap beam lberror (%)
in. (N-m)
fy = 68 ksi (469
MPa)
fy = 75 ksi (517
MPa)
1*10^8 lb-in.
(113*10^5 N-m)
1.5*10^8 lb-in.
(17*10^6 N-m)
5*10^8 lb-in.
(56.5*10^6 N-m)
1*10^8 lb-in.
(113*10^5 N-m)
5*10^8 lb-in.
(56.5*10^6 N-m)
177
Phase
error
(%)
Total
error
(%)
41
20
61
24
23
47
11
33
44
41
20
61
9.4
29
38
Total
error
(%)
11
26
37
11
28
39
11
29
40
8.9
22
31
Stiffness
Amplitude
degredation
error (%)
variables
wc=0.5,
wt=0
wc=0.8,
wt=0
wc=1,
wt=0.8
wc=0.8,
wt=1
Amplitude
error (%)
with Slip
without
Slip
Phase
error
(%)
Total
error
(%)
13
31
43
8.9
22
30
SHAKE TABLE
Abutment 1
Bent 1
SHAKE TABLE
Bent 2
SHAKE TABLE
Bent 3
178
Abutment 2
(a)
(b)
Fig. 2 Close view of (a) bottom and (b) top grid systems
DT7
3ETR3
57
56 58
50
49 51
59
52
62
6061
55
5354
43
42 44
45
48
4647
3ETR5
3ETR6
36
35 37
29
28 30
212223
41
3940
34
31 3233
24 252627
38
1617 18
19 20
1112
13
1415
6 7
9 10
1 2
4 5
3EBR5
3EBR7
3ETR4
3EBR6
3EBR8
Fig. 3 Grid system and LVDT locations on column in the Bent 3 east column
179
Lateral Component
0.3
Acceleration (g)
0.2
0.1
0
0
10
20
30
40
-0.1
-0.2
-0.3
Time (sec)
Longitudinal Component
0.25
.
0.2
Acceleration (g)
0.15
0.1
0.05
0
-0.05 0
10
20
30
-0.1
-0.15
-0.2
Time (sec)
180
40
m]
in
75.24
14
6 .8
. [3
i
.9 1
50 ]
m
.7 3
75.
.2
. [1
9m
6 7 .8
RC
6 in .
[1 .7 2
m]
24
Tower
2 88
LC
103.8
6
in. [2
.64 m
]
3 6 in
. [ 0 .9
1
in. [7
.32m
28
8
.[
7.
32
.5 3
in
29
m
]
Bent 3
Shake Table
LC Location of Left Camera
RC Location of Right Camera
Fig. 7 Location of the Bent 3 east column and the aluminum tower
181
Unconfined
Stress (ksi)
5
4
3
2
1
0
0
0.02
0.04
0.06
0.08
0.1
0.12
Strain (in./in.)
Tension Model
Bhide_87
Vecchio_82
Collins_87
350
300
Stress (psi)
250
200
150
100
50
0
0
0.005
0.01
Strain
182
0.015
0.02
Steel Model
100000
90000
80000
Stress (psi)
70000
Static
60000
Dynamic
50000
40000
30000
20000
10000
0
0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
Strain (in./in.)
Stress (psi)
40000
20000
0
-0.01
-20000
0.01
0.02
0.03
0.04
0.05
-40000
-60000
Ma et al. (1976)
-80000
FE
-100000
Strain (in./in.)
183
Force (kips)
15
3 in.
10
Curve
1 in
5
0.5 in
0
0
-5
Displacement (in.)
Fig. 12 Load-deflection curves for monotonic loading analysis of FE model of the bridge
column with various mesh densities
2.5
Photogrammetry
DT7 & (DS5+DS1) combined
components
Displacement (in.)
1.5
0.5
0
0
10
15
20
25
30
35
40
-0.5
-1
-1.5
Time (sec)
184
Lateral Displacement
Photogrammetry #2
1.00
FE
0.50
Displacement (in.)
0.00
0
10
15
20
25
30
35
40
-0.50
-1.00
-1.50
-2.00
Time (sec)
1.50
Displacement (in.)
FE
1.00
0.50
0.00
5
10
15
20
25
30
35
40
-0.50
-1.00
-1.50
-2.00
Time (sec)
185
FE @5in top
0.1
0.08
D isp. (in.)
0.06
0.04
0.02
0
-0.02 5
10
15
20
25
30
35
40
-0.04
-0.06
-0.08
Time (sec)
0.1
0.08
0.06
D isp. (in.)
0.04
0.02
0
-0.02
10
15
20
25
30
35
-0.04
-0.06
-0.08
Time (sec)
186
40
0.08
0.06
Disp. (in.)
0.04
0.02
0
-0.02
10
15
20
25
30
35
40
-0.04
-0.06
-0.08
Time (sec)
0.08
FE @5 in top
0.06
Disp. (in.)
0.04
0.02
0
-0.02 5
10
15
20
25
30
35
40
-0.04
-0.06
-0.08
Time (sec)
FE @5 in.
0.01
LVDT @ 5 in.
0.005
Rotation (rad)
0
5
10
15
20
25
30
35
40
-0.005
-0.01
-0.015
-0.02
-0.025
Time (sec)
Fig. 20 The cross-sectional rotation comparison between FE analysis result and LVDT
measurements at the location of LVDT 3ETR3 and 3ETR4 (fy=64 ksi)*
0.01
FE@10
LVDT@12
0.005
Rotation (rad)
0
5
10
15
20
25
30
35
40
-0.005
-0.01
-0.015
-0.02
-0.025
Time (sec)
Fig. 21 The cross-sectional rotation comparison between FE analysis result and LVDT
measurements at the location of LVDT 3ETR5 and 3ETR6 (fy=64 ksi) *
*
FE @5 in.
0.01
LVDT @ 5 in.
0.005
Rotation (rad)
0
5
10
15
20
25
30
35
40
-0.005
-0.01
-0.015
-0.02
-0.025
Time (sec)
Fig. 22 The cross-sectional rotation comparison between FE analysis result and LVDT
measurements at the location of LVDT 3ETR3 and 3ETR4 (fy=70 ksi) *
FE@10
0.01
LVDT@12
0.005
Rotation (rad)
0
5
10
15
20
25
30
35
40
-0.005
-0.01
-0.015
-0.02
-0.025
Time (sec)
Fig. 23 The cross-sectional rotation comparison between FE analysis result and LVDT
measurements at the location of LVDT 3ETR5 and 3ETR6 (fy=70 ksi) *
*
189
FE @5 in.
0.01
LVDT @ 5 in.
0.005
0
Rotation (rad)
10
15
20
25
30
35
40
-0.005
-0.01
-0.015
-0.02
-0.025
Time (sec)
Fig. 24 The cross-sectional rotation comparison between FE analysis result and LVDT
measurements at the location of LVDT 3ETR3 and 3ETR4 (fy=75 ksi) *
FE@10
0.01
LVDT@12
0.005
Rotation (rad)
0
5
10
15
20
25
30
35
40
-0.005
-0.01
-0.015
-0.02
-0.025
Time (sec)
Fig. 25 The cross-sectional rotation comparison between FE analysis result and LVDT
measurements at the location of LVDT 3ETR5 and 3ETR6 (fy=75 ksi) *
*
FE @5 in.
0.01
LVDT @ 5 in.
0.005
Rotation (rad)
0
5
10
15
20
25
30
35
40
-0.005
-0.01
-0.015
-0.02
Time (sec)
Fig. 26 The cross-sectional rotation comparison between FE analysis result and LVDT
measurements at the location of LVDT 3ETR3 and 3ETR4 (fy=80 ksi) *
FE @10 in.
0.01
LVDT @ 12
in.
0.005
0
Rotation (rad)
10
15
20
25
30
35
40
-0.005
-0.01
-0.015
-0.02
-0.025
Time (sec)
Fig. 27 The cross-sectional rotation comparison between FE analysis result and LVDT
measurements at the location of LVDT 3ETR5 and 3ETR6 (fy=80 ksi) *
*
Amplitude Error
35
Phase Error
30
Total Error
Error (%)
25
20
15
10
5
0
64
70
75
80
Fig. 28 The comparison of FDE index errors for different yield strength of steel (FE vs.
LVDT vertical displacement measurements)
Cross-sectional Rotation @ 5 in.
Amplitude Error
Phase Error
Total Error
35
30
Error (%)
25
20
15
10
5
0
64
70
75
80
Fig. 29 The comparison of FDE index errors for cross-sectional rotation at 5 in. below
cap beam using different yield strength of steel (FE vs. LVDT vertical displacement
measurements)
192
Phase Error
Total Error
30
Error (%)
25
20
15
10
5
0
64
70
75
80
Fig. 30 The comparison of FDE index errors for cross-sectional rotation at 12 in. below
cap beam using different yield strength of steel (FE vs. LVDT vertical displacement
measurements)
Vertical Disp. @Column Face
LVDT 3ETR4
Disp. (in.)
0.06
FE @5in top
0.04
0.02
0
-0.02
10
15
20
25
30
35
40
-0.04
-0.06
-0.08
Time (sec)
Disp. (in.)
0.06
FE @5in top
0.04
0.02
0
0
10
15
20
25
30
35
40
-0.02
-0.04
-0.06
-0.08
Time (sec)
0.15
.
FE @5in top
Disp. (in.)
0.1
0.05
0
0
10
15
20
25
30
35
40
-0.05
-0.1
Time (sec)
Linear-elastic rotational spring with 10^8 lb-in. capacity, wc =0.8 and wt =1.0.
194
0.06
.
FE @5in top
Disp. (in.)
0.04
0.02
0
-0.02
10
15
20
25
30
35
40
-0.04
-0.06
-0.08
Time (sec)
LVDT 3ETR4
FE @5in top
0.12
0.1
Disp. (in.)
0.08
0.06
0.04
0.02
0
-0.02 0
10
15
20
25
30
35
40
-0.04
-0.06
-0.08
Time (sec)
195
70
Total error
60
Error (%)
50
40
30
20
10
0
1*10^8 lb-in. (113*10^5
N-m)
Fig. 36 The comparison of FDE index errors for different stiffness of the cap beam (FE
vs. LVDT vertical displacement measurements) (fy=68 ksi) *
196
70
Total error
60
Error (%)
50
40
30
20
10
0
1*10^8 lb-in. (113*10^5 N-m)
Fig. 37 The comparison of FDE index errors for different stiffness of the cap beam (FE
vs. LVDT vertical displacement measurements) (fy=75 ksi) *
1-dc
0.8
0.6
0.4
0.2
0
0
0.02
0.04
0.06
0.08
0.1
0.12
Fig. 38 The compression stiffness degradation model for core concrete of the bridge
column
Cover Concrete Damage Degredation Model
1.2
1
1-dc
0.8
0.6
0.4
0.2
0
0
0.005
0.01
0.015
0.02
Fig. 39 The compression stiffness degradation model for cover concrete of the bridge
column
198
1-dt
0.8
0.6
0.4
0.2
0
0.00
0.01
0.02
0.03
0.04
0.05
0.06
0.07
Fig. 40 The tension stiffness degradation model for core and cover concrete of the bridge
column
LVDT 3ETR4
FE @5in top
0.15
D isp. (in.)
0.1
0.05
0
5
10
15
20
25
30
35
-0.05
-0.1
Time (sec)
199
40
FE @5in top
D isp . (in .)
0.1
0.05
0
5
10
15
20
25
30
35
40
-0.05
-0.1
Time (sec)
0.15
Disp. (in.)
0.1
0.05
0
5
10
15
20
25
30
35
-0.05
-0.1
Time (sec)
200
40
0.12
FE @5in top
0.1
0.08
Disp. (in.)
0.06
0.04
0.02
0
-0.02 5
10
15
20
25
30
35
40
-0.04
-0.06
-0.08
Time (sec)
50
Amplitude error
Phase error
45
Total error
40
Error (%)
35
30
25
20
15
10
5
0
wc=0.5 wt=0
wc=0.8 wt=0
wc=1 wt=0.8
wc=0.8 wt=1
Fig. 45 The comparison of FDE index errors for different recovery and degradation
parameters (FE vs. LVDT vertical displacement measurements) *
*
12000
Force (lb)
10000
8000
6000
4000
2000
0
0
0.02
0.04
0.06
0.08
0.1
Slip (in.)
202
0.2
Disp. (in.)
0.15
0.1
0.05
0
5
10
15
20
25
30
35
40
-0.05
-0.1
Time (sec)
FE @5 in.
0.01
LVDT @ 5 in.
0.005
Rotation (rad)
0
5
10
15
20
25
30
35
40
-0.005
-0.01
-0.015
-0.02
-0.025
Time (sec)
Fig. 49 The cross-sectional rotation comparison at the location of LVDT 3ETR3 and
3ETR4*
*
Infinitely stiff cap beam, bond-slip effect included at the interfaces, wc=0.8, wt= 1.
203
0.015
FE@10
LVDT@12
0.01
Rotation (rad)
0.005
0
5
10
15
20
25
30
35
40
-0.005
-0.01
-0.015
-0.02
-0.025
Time (sec)
Fig. 50 The cross-sectional rotation comparison at the location of LVDT 3ETR5 and
3ETR6*
Yield Strength= 64 ksi (441 Mpa)
Amplitude error
Phase error
70
Total error
60
Error (%)
50
40
30
20
10
0
with Slip
without Slip
Fig. 51 The comparison of FDE index errors for bond-slip models (FE vs. LVDT vertical
displacement measurements)*
*
Infinitely stiff cap beam, bond-slip effect included at the interfaces, wc=0.8, wt= 1.
204
Right Line_Photogrammetry
Right Line_FEA
50
40
30
20
10
0
-0.2
-0.1
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
Fig. 52 Deformed shape of the bridge column at the maximum drift during Test 4D
(fy=64 ksi)*
Infinitely stiff cap beam, bond-slip effect included at the interfaces, wc=0.8, wt= 1.
205
Middle
Line_Photogrammetry
Middle Line_FEA
Left
Line_Photogrammetry
Left Line_FEA
60
Right
Line_Photogrammetry
Right Line_FEA
70
50
40
30
20
10
0
-0.2
-0.1
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
Fig. 53 Deformed shape of the bridge column at the maximum drift during Test 4D
(fy=80 ksi) *
Infinitely stiff cap beam, bond-slip effect included at the interfaces, wc=0.8, wt= 1.
206
REFERENCES
1. Baker A.L.L. Ultimate load theory applied to the design of reinforced and
prestressed concrete frames. Concrete Publications Ltd., London. 1956.
2. Baker A.L.L., Amarakone A.M.N. Inelastic hyperstatic frame analysis. Proc.
International Symposium on the Flexural Mechanics of Reinforced Concrete,
ACI SP-12 Miami, pp. 85-142. 1964.
3. Mattock A.H. Rotational capacity of hinging regions in reinforced concrete
beams. Proc. International Symposium on the Flexural Mechanics of Reinforced
Concrete, ACI SP-12 Miami, pp. 143-181. 1964.
4. Mattock, A.H. Discussion of rotational capacity of hinging regions in reinforced
concrete beams. by Corley, W.G., Journal of the Structural Division, ASCE, V.
93, No. ST2, Apr., pp. 519-522. 1967.
5. Corley W. G. Rotational capacity of reinforced concrete beams. J Struct Division
ASCE 1966; V. 92, No. ST5, pp. 121-146.
6. Park R., Priestley M. J. N., Gill W. D. Ductility of square-confined concrete
columns. J Struct ASCE 1982; 108 No. ST4. 929-950.
7. Priestley M. J. N., Park R. Strength and ductility of concrete bridge columns
under seismic loading. ACI Struct J 1987 V.84 Issue 1. pp. 61-76.
8. Paulay T., Priestley M. J. N. Seismic design of reinforced concrete and masonry
structures. John Wiley & Sons, Inc. 1992.
9. Sheikh S.A., Khoury S.S. Confined concrete columns with stubs. ACI Struct J
1993 V. 90 No. 4 pp. 414-431.
207
10. Mendis P. Plastic hinge lengths of normal and high-strength concrete in flexure.
Advances in Structural Engineering 2001 V. 4 No. 4 pp. 189-195.
11. Firat
Alemdar
Z.,
Browning
J.,
Olafsen
J.
2010.
Photogrammetric
M.
Design
of
Earthquake-Resistant
York:McGraw-Hill, 1986.
208
Buildings.
New
20. Hosoya, H. et al. Strain rate in the members of reinforced concrete frame
structure during an earthquake and its effects. Journal of Structural and
Construction Engineering (Transactions of AIJ) 499 (1997): 77-83.
21. Mahin, S.A., V.V. Bertero, M.B. Atalya and D. Rea. Rate of loading effects on
uncracked and repaired reinforced concrete members. Report UCB/EERC-72/09.
Berkeley: EERC, University of California, 1972.
22. Bhide, S. B., and Collins, M. P. Reinforced concrete elements in shear and
tension. Publication No. 87-02, Department of Civil Engineering, University of
Toronto, Jan. 1987, 308 pp.
23. Vecchio, F. J., and Collins, M. P. Response of reinforced concrete to in-plane
shear and normal stresses. Publ. No. 82-03, Dept. of Civil Engineering, Univ. of
Toronto. 1982.
24. Collins, M. P., and Mitchell, D.
field
theory.
Version
http://www.ecf.utoronto.ca/~bentz/r2k.htm.
209
1.0.5,
last
accessed
2010,
29. Dragovich JJ., Lepage A. FDE index for goodness-of-fit between measured and
calculated response signals. Earthquake Engineering and Structural Dynamics
2009;38:1751-1758.
30. Sezen, H. Seismic Behavior and Modeling of Reinforced Concrete Building
Columns. Ph.D. Thesis. University of California, Berkeley. 2002.
210
1147 Learned Hall, Department of Civil, Env, and Arch. Engineering, University of Kansas, Kansas,
USA
2150 Learned Hall, Department of Civil, Env, and Arch. Engineering, University of Kansas, Kansas,
USA
2150 Learned Hall, Department of Civil, Env, and Arch. Engineering, University of Kansas, Kansas,
USA
Corresponding author Tel: +1 785 864 3760 Fax: +1 785 864 5631 E-mail addresses:
zeynep@ku.edu
(Z. F. Alemdar).
1. Introduction
The determination of magnitude and location of inelastic deformations in
reinforced concrete bridge columns is a critical step for characterizing the performance
of the bridge system in earthquake events. Although it is possible to believe that some
ductility will be provided by beam hinges in bridge systems, it is generally the columns
of the bridges that must have inelastic rotational capacity. Bridge systems are designed to
keep inelastic behavior in the columns and away from the superstructure, which is
different than building systems. Plastic hinging regions indicate the area of concentrated
damage for bridge columns that experience inelastic deformations. Therefore, a
consistent prediction of a plastic hinge length is also necessary to evaluate the length of
the column that needs to be confined along the critical section. Simple expressions for
211
plastic hinging regions also enable the development of estimates of column drift capacity
and drift demand.
The plastic hinge length in reinforced concrete columns has been studied widely
by many researchers [2-16]. Previous experimental research has generally been
conducted using static loading and for small-scale components such as beams and
columns. The real load distribution due to earthquake excitations of a structure would be
different from static loading, and elements would behave differently. In addition, an
evaluation of large-scale system behavior will include the effects of moment
redistribution and progression of yielding throughout the entire structure.
Only a few of the previous studies have focused on determining the plastic hinge
length in reinforced concrete bridge columns, and instead have focused on building-type
columns. The location and the progression of plastic hinge length in reinforced concrete
bridge columns would differ from reinforced concrete buildings. In addition, many of the
parameters such as axial load ratio, diameter of the longitudinal reinforcement and the
cross-sectional dimensions of the columns in the previous tests are not appropriate for
reinforced concrete bridge columns [12-16, 24].
Expressions that have been developed to estimate the length of plastic hinges
have either been based on the maximum drift at the top of the column, or the spread of
plastic behavior in the hinging regions. Plastic hinge lengths have been calculated for 51
of the reported tests from the measured maximum top displacement using the lumpedplasticity model [7, 10, 17, 20, 21, 26-29]. Only 21 test data included the measured
plastic hinge length from using the spread of plasticity approach [15, 16, 19, 22, 23, and
25]. The lumped-plasticity model has generally been used to simulate the nonlinear
response of bridge columns and estimate the ultimate displacement capacity, which is an
212
213
displacements at the cap beam on the bridge columns of Bent 3 and compared with the
calculated response from the expression [30].
2.
from static tests. A suite of 115 column tests that represent typical elements in reinforced
concrete buildings and bridges that encompass a wide range of parameters were used.
Axial load level, compressive strength of concrete, yield strength of longitudinal
reinforcement, diameter of longitudinal reinforcement, and section dimensions are some
of the main variables in the static tests. The minimum and maximum ranges of the main
variables used in the literature are given in Table 1.
In the studies performed before 1983, the cross-section dimensions of a column
were the main parameters used to determine the plastic hinge length of a column. With
the results of Manders tests at the University of Canterbury [8], the strain penetration
effect of the longitudinal reinforcement was found to be an important factor within the
plastic hinge region, and the diameter of the longitudinal reinforcement was incorporated
into new expressions. After the effect of axial load on plastic hinge length calculation
was discovered, Soesianawati et al. (1986) and Watson et al. (1994) added the axial load
ratio into their expressions. Later expressions incorporate dimensionless parameters such
as longitudinal reinforcement ratio, and shear span to depth ratio into expressions [15].
Table 2 shows various plastic hinge length expressions that have been developed to
predict the plastic hinge length of reinforced concrete columns.
A small suite of 10 dynamic test results with relevance to determining plastic
hinge length expressions have been collected. Hachem et al. (2003) and Dodd et al.
214
(2000) used the equation proposed by Priestley et al. (1992) and estimated the plastic
hinge length before testing circular bridge columns under uniaxial and bidirectional
earthquake loading. The main parameters in the tests were aspect ratio of the columns (4
to 10) to characterize the flexurally dominated columns, axial load level (0.04Agfc 0.4
Agfc), and longitudinal reinforcement ratio (1.2%-1.62%). The plastic hinge length was
measured according to the spread of plasticity model in the study of Hachem et al.
(2003). Dodd et al. (2000) calculated the plastic hinge lengths from the maximum drift of
the columns. After testing the columns, it was also observed that the estimated plastic
hinge lengths were within the range of measured values as shown in Table 3. The
average measured plastic hinge length value from the testing of four identical columns
[19] was about 13 in. (330 mm) as given in PEER 2003/06 report. It was clear from the
experimental values that the equation proposed by Priestley et al. (1992) overestimated
the plastic hinge length for all of the columns as shown in Table 3.
Two new expressions to calculate the plastic hinge length in reinforced concrete
bridge columns have been derived using the measured plastic hinge length values from
yield curvature and the plastic hinge length values that were calibrated to calculate the
correct maximum top displacement during the static and dynamic tests. The experimental
data was limited to contain the data which characterize modern bridge column design
criteria as described next, and included 72 out of the 115 tests from the literature. Fifty
one plastic hinge length data included the plastic hinge length values calculated from the
measured top displacement as given in Table 4. The remaining 21 test data measured the
plastic hinge length as defined where the curvature is bigger than the yield curvature of
the columns that were tested, as shown in Table 5. The test data was limited to represent
modern bridge columns, which meet the following design criteria:
215
An axial load ratio P/fcAg less than or equal to 0.3, where P is the column
axial load, fc is the concrete compressive strength, Ag is the gross section
area of the column,
Each parameter of the equations in Table 1.2 was considered in the analysis to
determine their affect on estimates of plastic hinge lengths. In order to obtain the best fit
expressions for the plastic hinge length over the cross-sectional dimension, multivariable
linear regression analysis was used. A number of different combinations of parameters
were examined including the length of the columns, axial load ratio, diameter of the
longitudinal reinforcement, yield strength of the reinforcement, compressive strength of
the concrete, shear span to depth ratio, and the diameter of the longitudinal reinforcement
normalized with yield strength of the reinforcement and the compressive strength of the
concrete. The determining factor for selecting the best-fit equation was the minimum
coefficient of variation (COV) of measured Lp/d (ratio of plastic hinge length to section
dimension) to that of calculated Lp/d. After analyses of the various combinations, the
column height and the diameter of the longitudinal reinforcement normalized with yield
strength of the reinforcement and the compressive strength of the concrete (representing
bending along the column height and the bond-slip deformation effect) were found to be
the parameters that produced the best-fit expressions. The best-fit expression derived
using 51 test data based on calibrated value of lp to calculate the maximum drift capacity
is given in Eq. 1.
216
Lp
d
Lp
d
1 3 f y db
L
+
+
5 500 f c' 1000
(1a)
3 f y db
1
L
+
+
'
4 10000 f c 25000
(1b)
The best-fit expression based on measured spread of plasticity (21 test data set) is given
in Eq. 2.
Lp
d
Lp
d
f y db
3
L
+
+
(psi and in.)
10 1000 f c' 2500
(2a)
3 f y db
3
L
+
+
10 100000 f c' 5000
(2b)
The new expressions were compared with previous plastic hinge length models.
In order to compare the different expressions, the measured plastic hinge length was
divided by the calculated plastic hinge length obtained from the expressions available.
The best fit lines were drawn using the ratios of measured to calculated plastic hinge
lengths based on each equation available in the literature. Figure 2 shows the
comparisons of the ratio of measured to calculated plastic hinge lengths with varied
parameter (fydb/fc) in Eq. 1 for all data. The ratio of measured to calculated plastic
hinge length results were closest to unity for the new expression. The ratio of measured
to calculated plastic hinge length values determined by the equations of Park (1982) and
Paulay (1992) were the next best for representing the data at low axial loads. The
comparison of the ratio of measured to calculated plastic hinge length result with varied
bridge column height is shown in Fig. 3 for Eq. 1. The ratio of measured to calculated
plastic hinge length results from the new expression (Eq. 1) were the closest to unity
with respect to the column height. The ratio of measured to calculated plastic hinge
length values determined by the equations of Mander (1983), Park (1987), and Paulay
217
(1992) intersected unity well, which was very similar to the results found in Fig. 1 except
Paulay (1992) overestimated the plastic hinge length with respect to column height.
The calculated COV values for the ratio of measured to calculated plastic hinge
length are between 0.37 and 0.57 for each expression. The new expression had the lowest
COV value equal to 0.37. The COV of Park et al. (1982) equation was 0.39, which is the
second best COV value after the new expression. Baes equation gave the highest COV
(0.57) for all data with low axial load. The COV values were 0.41, 0.42, 0.45, and 0.45
for Paulay (1992), Park (1987), Mander (1983), and Berry (2008) respectively.
Figure 4 shows the comparisons of the ratio of measured to calculated plastic
hinge lengths with varied value of the first parameter in Equation 2 and the best fit lines
drawn using the ratios for each available expression based on the 21 test data set. The
ratio of measured to calculated plastic hinge length results were closest to unity for the
new expression with respect to fydb/fc. The ratio of measured to calculated plastic hinge
length values determined by the equations of Mander (1983) and Park (1987) were the
next best results after the new expression values. The comparison of the ratio of
measured to calculated plastic hinge lengths with varied bridge column height is shown
in Fig. 5 for the 21 test data set. The ratio of measured to calculated plastic hinge length
results from the new expression (Eq. 2) were the closest to unity with respect to the
column height. The ratio of measured to calculated plastic hinge length values
determined by the equations of Mander (1983) and Park (1987) followed a similar trend
as the new expression.
The calculated COV values for each expression using the analysis of 21 test data
based on the measured spread of plasticity are between 0.27 and 0.46. The new
expression had a COV value equal to 0.30, and the COV of the Mander (1983) equation
218
was 0.27. The Bae (2005) and Berry (2008) equations gave the highest COV (0.46) for
low axial load level. The COV values were 0.32, 0.40, and 0.44 for Park (1987), Park
(1982), and Paulay (1992) respectively.
The COV of Eq. 1 was also determined using the 21 test data of plastic hinge
length derived considering the length of the column with curvature values greater than
the yield curvature (>y) to determine how reliable Eq. 1 is to estimate the actual spread
of plasticity. The ratios of plastic hinge length values from the new expression (Eq. 1) to
the measured plastic hinge length results were calculated. The COV was found to be
0.34, which is an average COV compared to that of the other expressions. This analysis
shows that the behavior of columns when considering the maximum drift capacity and
spread of plasticity (>y) is different.
It is of interest to consider the effect of the axial load level when using each
expression. All the expressions except Berry (2008), which was derived only for bridge
columns with P/Agfc less than or equal to 0.3, were developed for a wide range of axial
loads. In the literature, some of the plastic hinge length expressions work well for high
axial load ratios, whereas they fail to give satisfactory results for low axial load ratios.
Figure 6 shows the effectiveness of the plastic hinge length calculations for a wide range
of axial load ratios. Although only Baes equation includes an axial load parameter, some
of the other equations show considerable changes of measured to calculated ratios of
Lp/d with increasing levels of axial load. It is important to note that the COV values
varied between 0.16 and 0.30 for high axial load ratios. The COV values were 0.16, 0.17,
0.17, 0.18, 0.27, 0.28, 0.28, and 0.30 for Park (1987), Mander (1983), Berry (2008),
Priestley (1992), Bae (2005), Park et al. (1982), and the proposed equations (1.1 and 1.2)
respectively. It appears that most equations have a much better correlation for the plastic
219
hinge length with high axial load, and the behavior of columns is quite different under
high axial load than low axial load. The proposed expression had a good COV value
equal to 0.30, which compared very well to the COV at low axial load ratio of (0.37).
Therefore, the use of Eq. 1 and Eq. 2 provide satisfactory predictions both for high and
low axial load levels. At high axial load ratios, the proposed expression can be
normalized by 1.4 to provide a good estimate for the plastic hinge length.
3.
column tested at UNR as described previously. A drift expression was derived using a
lumped-plasticity model with a fixed plastic hinge length (lp) for a double curvature
column. Flexural deformations and slip of the longitudinal reinforcement at the cap beam
and footing interface were considered in the maximum drift expression, and given in
Equation 1.3. The first and last terms in Eq. 3 represent the flexural displacement in the
elastic and plastic range, and the second term represents the slip contribution to the top
displacement.
lp
d
yl 2
6
db f y
48 E s
*
f 'c
L
+ pl p (L l p )
d d'
(3)
The bridge column was tested using the biaxial applications of scaled motions
recorded at the Century City Country Club during the 1994 Northridge, California
earthquake [31]. The maximum drift was measured to be 5.3 in. in the transverse
direction of the bridge after all the motions were applied. The maximum drift values of
the bridge column were calculated using the plastic hinge length expressions (Table 2)
and compared with the measured data in Table 5. The Park (1982) plastic hinge length
220
equation gave the closest value (5.94 in.) to the measured test data. The proposed
expression (Eq. 1) also estimated the measured data very well and provided the second
best estimate of 6.5 in. after the Park (1982) equation. The maximum drift value was also
determined using Equation 2 and found to be 7.1 in., which is the third best correlation to
the real test data.
4.
Conclusions
A review of literature shows that an expression for plastic hinge length can be
developed based on values used to produce best estimates of maximum drift capacity or
can be developed to represent the spread of plasticity along the height of the column.
This study produced equations that characterize the behavior of these two groups of data
better than any existing equation of lp, and shows that the behavior is different between
these two methods. The following conclusions are made:
1. Estimates of plastic hinge length used to calculate the maximum drift
capacity are shorter in value than estimates of plastic hinge length
based on spread of plastic curvature over height of column.
2. The spread of plasticity at low axial loads is different than at high
axial loads. Most existing equations are better fit at high axial loads.
The proposed equations have nearly constant fit across axial loads,
but best fit at low axial loads, which represent bridge columns.
3. The maximum drift of the one column tested under dynamic loads
was best represented using the simple plastic hinge length estimate
proposed by Park (1982). More parametric study is needed to
understand the spread of plasticity.
221
db, in.
fc, ksi
fy, ksi
2
P/Pu
(mm)
(mm)
(%)
(kN/mm )
(kN/mm2)
(1)
(2)
(3)
(4)
(5)
(6)
0.24 (6.1)
0.7
3.1 (0.021)
43.5 (0.3)
16.25
75.55
(0.11)
(0.52)
Minimum
0.003 25 (635)
Maximum
0.77
360
1.75
(9144)
(44.45)
5.65
Expressions
0.4*d (1)
0.08*L+.15dblfye (4)>0.3dblfye
(0.08*L+.022dblfye>0.3dblfye)
6.35*(db)+0.06*L
(32*(db)+0.06*L)
(5)
{(0.3*(P/P0 )+3*(As (6)/Ag)0.1)*(L/h)+.25}*h
222
Hachem
(2003)
Dodd
(2000)
L/d(1)
Calculated
Lp, in. (mm)
Measured
Lp, in.
(mm)
13.30 (338)
13 (330)
4.84 (123)
3.54 (90)
4.84 (123)
3.35 (85)
6.73 (171)
4.92 (125)
6.73 (171)
5.31 (135)
10
8.62 (219)
7.68 (195)
10
8.62 (219)
5.91 (150)
4.84 (123)
3.35 (85)
6.73 (171)
5.31 (135)
4.84 (123)
3.74 (95)
1: Aspect ratio
223
Park
(1982)
Priestley
(1987)
Tanaka
(1990)
Kovacic
(1995)
d or
h, in.
(mm)
(1)
22
(559)
22
(559)
19.69
(500)
19.69
(500)
19.69
(500)
23.62
(600)
15.75
(400)
15.75
(400)
29.53
(750)
29.53
(750)
29.53
(750)
15.75
(400)
15.75
(400)
15.75
(400)
15.75
(400)
21.65
(550)
21.65
(550)
21.65
(550)
21.65
(550)
6.46
(164)
6.46
(164)
6.46
(164)
6.46
(164)
Measured
Lp, in.
(mm)
(2)
9.68
(246)
7.48
(190)
10.63
(270)
11.42
(290)
8.86
(225)
8.27
(210)
8.51
(216)
8.66
(220)
10.93
(278)
11.81
(300)
12.31
(313)
6.77
(172)
Lp/d
(3)
P/f'cAg
(4)
L/d
(5)
0.44
0.26
2.18
0.34
0.21
2.18
0.54
0.11
4.00
0.58
0.11
4.00
0.45
0.00
5.50
0.35
0.15
2.00
0.54
0.20
4.00
0.55
0.21
4.00
0.37
0.10
4.27
0.40
0.30
4.27
0.42
0.30
4.27
0.43
0.20
4.00
0.55
0.20
4.00
0.67
0.20
4.00
0.70
0.20
4.00
0.39
0.10
3.00
0.63
0.10
3.00
0.67
0.30
3.00
0.86
0.30
3.00
1.5 (38)
0.25
0.16
8.46
2.76 (70)
0.47
0.17
8.46
3.74 (95)
0.63
0.14
8.46
1.77 (45)
0.30
0.19
8.46
8.7 (221)
10.6
(269)
11.06
(281)
8.35
(212)
13.66
(347)
14.45
(367)
18.66
(474)
224
L, in.
(mm)
(6)
48
(1219)
48
(1219)
70
(1778)
108
(2743)
108
(2743)
47
(1194)
63
(1600)
63
(1600)
126
(3200)
126
(3200)
126
(3200)
63
(1600)
63
(1600)
63
(1600)
63
(1600)
65
(1651)
65
(1651)
65
(1651)
65
(1651)
50
(1270)
50
(1270)
50
(1270)
50
(1270)
db,
in.
(mm)
(7)
0.94
(24)
0.94
(24)
0.51
(13)
0.51
(13)
0.51
(13)
0.94
(24)
0.63
(16)
0.63
(16)
0.39
(10)
0.39
(10)
0.39
(10)
0.79
(20)
0.79
(20)
0.79
(20)
0.79
(20)
0.79
(20)
0.5
(13)
0.5
(13)
0.5
(13)
0.47
(12)
0.47
(12)
0.47
(12)
0.47
(12)
fy, ksi
(MPa)
(8)
55.1
(380)
55.1
(380)
54.08
(373)
53.8
(371)
44.22
(305)
43.94
(303)
44.67
(308)
61.93
(427)
48.58
(335)
48.58
(335)
48.58
(335)
68.73
(474)
74.1
(511)
74.1
(511)
74.1
(511)
74.1
(511)
68
(469)
68
(469)
68
(469)
58
(400)
58
(400)
58
(400)
58
(400)
f'c,
ksi
(MPa)
(9)
3.35
(23)
6
(41)
4.81
(33)
5.04
(35)
5.8
(40)
4.35
(30)
3.77
(26)
3.63
(25)
5.8
(40)
4.21
(29)
4.64
(32)
3.71
(26)
4.64
(32)
4.64
(32)
4.65
(32)
4.65
(32)
l
(%)
(10)
1.79
1.79
2.70
2.70
2.57
2.43
2.43
1.51
2.84
2.84
2.84
1.57
1.57
1.57
1.57
1.25
6 (41)
1.25
6.15
(42)
1.25
6 (41)
1.25
4.82
(33)
4.74
(33)
5.61
(39)
8.43
(58)
5.65
5.65
5.65
5.65
Table 4 Column properties and calculated plastic hinge length values (continued)
Kovacic
(1995)
Lehman
and
Moehle
(2000)
Calderone
(2000)
Dodd
(2000)
Moyer
and
Kowalsky
(2002)
d or
h, in.
(mm)
(1)
6.46
(164)
6.46
(164)
5.73
(146)
5.73
(146)
24
(610)
24
(610)
24
(610)
24
(610)
24
(610)
24
(610)
24
(610)
24
(610)
7.9
(201)
7. 9
(201)
7.9
(201)
7. 9
(201)
7.9
(201)
7. 9
(201)
7.9
(201)
7. 9
(201)
7.9
(201)
18
(457)
18
(457)
Measured
Lp, in.
(mm)
(2)
Lp/d
(3)
P/f'cAg
(4)
L/d
(5)
2.76 (70)
0.47
0.08
8.46
4.13
(105)
0.70
0.17
8.46
3.23 (82)
0.55
0.04
8.46
2.76 (70)
0.47
0.08
8.46
14 (356)
0.58
0.07
4.00
22 (559)
0.92
0.07
8.00
25 (635)
1.04
0.07
10.00
14 (356)
0.58
0.07
4.00
14 (356)
0.58
0.07
4.00
0.56
0.09
3.00
1.02
0.09
8.00
29 (737)
1.21
0.09
10.00
3.54 (90)
0.45
0.04
4.00
3.35 (85)
0.42
0.05
4.00
0.62
0.05
7.00
0.67
0.05
7.00
0.97
0.04
10.00
0.75
0.06
10.00
3.35 (85)
0.42
0.04
4.00
5.31
(135)
0.67
0.04
7.00
3.74 (95)
0.47
0.04
4.00
0.50
0.04
4.67
0.50
0.04
4.67
13.5
(343)
24.5
(622)
4.92
(125)
5.31
(135)
7.68
(195)
5.91
(150)
8.97
(228)
8.97
(228)
225
L, in.
(mm)
(6)
50
(1270)
50
(1270)
50
(1270)
50
(1270)
96
(2438)
192
(4877)
240
(6096)
96
(2438)
96
(2438)
72
(1829)
192
(4877)
240
(6096)
31.6
(803)
31.6
(803)
55.3
(1405)
55.3
(1405)
79
(2007)
79
(2007)
31.6
(803)
55.3
(1405)
31.6
(803)
84
(2134)
84
(2134)
db,
in.
(mm)
(7)
0.47
(12)
0.47
(12)
0.47
(12)
0.47
(12)
0.63
(16)
0.63
(16)
0.63
(16)
0.63
(16)
0.63
(16)
0.75
(19)
0.75
(19)
0.75
(19)
0.24
(6)
0.24
(6)
0.24
(6)
0.24
(6)
0.24
(6)
0.24
(6)
0.24
(6)
0.24
(6)
0.24
(6)
0.75
(19)
0.75
(19)
fy, ksi
(MPa)
(8)
58
(400)
58
(400)
58
(400)
58
(400)
60
(414)
60
(414)
60
(414)
60
(414)
60
(414)
60
(414)
60
(414)
60
(414)
65
(448)
65
(448)
65
(448)
65
(448)
65
(448)
65
(448)
65
(448)
65
(448)
65
(448)
60
(414)
60
(414)
f'c,
ksi
(MPa)
(9)
9.33
(64)
9.4
(65)
9.11
(63)
9.46
(65)
4.2
(29)
4.2
(29)
4.2
(29)
4.2
(29)
4.2
(29)
l
(%)
(10)
5.65
5.65
4.71
4.71
1.50
1.50
1.50
0.75
3.00
4 (28)
2.70
4 (28)
2.70
4 (28)
2.70
5.9
(41)
4.64
(32)
5 (34)
4.39
(30)
5.42
(37)
4.22
(29)
5.99
(41)
5.86
(40)
5.35
(37)
4.74
(33)
4.96
(34)
1.62
1.62
1.62
1.62
1.62
1.62
1.62
1.62
1.62
2.00
2.00
Table 4 Column properties and calculated plastic hinge length values (continued)
d or Measured
h, in.
Lp, in.
(mm)
(mm)
Lp/d P/f'cAg
(1)
(2)
(3)
(4)
18
8.97
Moyer
0.50
0.04
(457)
(228)
and
Kowalsky
18
8.97
0.50
0.04
(2002)
(457)
(228)
15.95
22.01
1.38
0.08
(405)
(559)
Phan
15.93
17.05
1.07
0.08
(2007)
(405)
(433)
16.05
17.01
1.06
0.08
(408)
(432)
1: Section dimension of the column tested.
L/d
(5)
4.67
4.67
4.50
4.50
4.50
L, in.
(mm)
(6)
84
(2134)
84
(2134)
72
(1829)
72
(1829)
72
(1829)
db,
in.
(mm)
(7)
0.75
(19)
0.75
(19)
0.75
(19)
0.75
(19)
0.75
(19)
fy, ksi
(MPa)
(8)
60
(414)
60
(414)
73.6
(507)
73.6
(507)
73.6
(507)
f'c,
ksi
(MPa)
(9)
4.6
(32)
4.92
(34)
4.7
(32)
4.72
(33)
4.81
(33)
l
(%)
(10)
2.00
2.00
2.00
2.20
2.00
Cheok
and
Stone
(1986)
d or
h, in.
(mm)
(1)
9.8
(249)
9.8
(249)
9.8
(249)
9.8
(249)
9.8
(249)
Measured
Lp, in.
(mm)
(2)
11.6
(295)
10.2
(259)
7.5
(191)
9.6
(244)
10.6
(269)
Lp/d
(3)
P/f'cAg
(4)
L/d
(5)
1.18
0.20
3.00
1.05
0.10
6.00
0.76
0.10
3.00
0.98
0.20
3.00
1.08
0.10
6.00
226
L, in.
(mm)
(6)
29.5
(749)
59
(1499)
29.5
(749)
29.5
(749)
59
(1499)
db,
in.
(mm)
(7)
1.75
(44)
1.75
(44)
1.75
(44)
1.75
(44)
1.75
(44)
fy, ksi
(MPa)
(8)
57
(393)
57
(393)
57
(393)
57
(393)
57
(393)
f'c,
ksi
(MPa)
(9)
3.35
(23)
3.68
(25)
3.55
(24)
3.53
(24)
3.37
(23)
l
(%)
(10)
1.45
0.70
1.45
1.45
0.70
Table 5 Column properties and measured plastic hinge length values (continued)
Wong et
al.
(1990)
Kunnath
(1997)
Bae
(2005)
Watson
(1994)
Hachem
(2003)
(11)
d or
h, in.
(mm)
(1)
Measured
Lp, in.
(mm)
(2)
15.75
(400)
9.84
(250)
12
(305)
12
(305)
12
(305)
12
(305)
12
(305)
12
(305)
12
(305)
12
(305)
24
(610)
24
(610)
15.75
(400)
15.75
(400)
7.09
(180)
7.12
(181)
5.91
(150)
5.90
(150)
7.87
(200)
7.68
(195)
8.66
(220)
7.87
(200)
11.76
(299)
11.28
(287)
4.09
(104)
5.2
(132)
15.75
(400)
L, in.
(mm)
(6)
db,
in.
(mm)
(7)
fy, ksi
(MPa)
(8)
f'c,
ksi
(MPa)
(9)
l
(%)
(10)
31.5
(800)
0.63
(16)
61.35
(423)
5.51
(38)
3.20
54
(1372)
54
(1372)
54
(1372)
54
(1372)
54
(1372)
54
(1372)
54
(1372)
54
(1372)
103.5
(2629)
103.5
(2629)
63
(1600)
63
(1600)
0.38
(10)
0.38
(10)
0.38
(10)
0.38
(10)
0.38
(10)
0.38
(10)
0.38
(10)
0.38
(10)
0.63
(16)
0.63
(16)
0.63
(16)
0.63
(16)
65
(448)
65
(448)
65
(448)
65
(448)
65
(448)
65
(448)
65
(448)
65
(448)
64.67
(446)
64.67
(446)
64.67
(446)
68.73
(474)
4.21
(29)
4.21
(29)
4.76
(33)
4.76
(33)
4.71
(32)
3.92
(27)
3.92
(27)
3.92
(27)
6.38
(44)
6.38
(44)
5.8
(40)
5.95
(41)
Lp/d
(3)
P/f'cAg
(4)
L/d
(5)
1.42
0.19
2.00
0.59
0.09
4.50
0.59
0.09
4.50
0.49
0.09
4.50
0.49
0.09
4.50
0.66
0.09
4.50
0.64
0.10
4.50
0.72
0.10
4.50
0.66
0.10
4.50
0.49
0.20
4.31
0.47
0.20
4.31
0.26
0.10
4.00
0.33
0.30
4.00
6.46
(164)
0.41
0.30
4.00
63
(1600)
0.63
(16)
68.73
(474)
5.8
(40)
1.51
15.75
(400)
4.09
(104)
0.26
0.30
4.00
63
(1600)
0.63
(16)
68.73
(474)
6.09
(42)
1.51
16
(406)
13
(330)
0.81
0.06
6.00
96
(2438)
0.5
(13)
60
(414)
5.7
(39)
1.20
227
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
1.25
1.25
1.51
1.51
Authors
Test data
Park (1982)
Mander (1983)
Park (1987)
Priestley (1992)
Bae (2005)
Berry (2008)
Eq. 1
Eq. 2
max (in.)
(transverse
dir.)
5.3
5.9
9.4
9.2
10.4
4.0
7.1
6.5
7.1
lp (in.)
4.8
8.2
8.0
9.4
3.0
5.9
5.4
5.9
(% difference)
12.0
77.0
73.0
97.0
-25.0
35.0
23.0
23.0
SHAKE TABLE
Abutment 1
SHAKE TABLE
Bent 1
SHAKE TABLE
Bent 2
Bent 3
Abutment 2
Alemdar Eq. 1
Bae
Park 82
5
Lp/d -measured/Lp/d-estimated
Park 87
Bae
4
Paulay 92
Berry
Mander 83
Berry 08
3
Park_82
2
Mander
1
Park_87
Alemdar Eq.1
0
0
100
200
300
400
500
fy *db/fc
600
700
Paulay
800
900
1000
Figure 2 Comparisons of the plastic hinge length expressions results with Equation 1
228
Alemdar Eq.
1
Bae
6
.
Park 82
Lp/d -measured/Lp/d-estimated
5
4
Park 87
Berry
Park_82
Bae
Paulay 92
Mander 83
Berry 08
Mander
2
Park_87
1
Alemdar Eq. 1
Paulay
0
50
100
150
200
250
300
L (in.)
Figure 3 Comparisons of the plastic hinge length expressions results with Equation 1
Lp/d -measured/Lp/d-estimated
Alemdar Eq. 2
Bae
Park_82
Park 82
Bae
Park 87
Berry
Paulay 92
4
Mander 83
Berry 08
3
2
Alemdar Eq. 2
1
0
0
200 Park_87
400
600 Paulay
800
Park_87
1000
fy *db/fc
1200
1400
1600
1800
Mander
2000
Figure 4 Comparisons of the plastic hinge length expressions results with Eq. 2
229
Alemdar Eq. 2
Bae
Lp/d -measured/Lp/d-estimated
Park_82
Park 82
Berry
Park 87
Bae
Paulay 92
Mander 83
Berry 08
1
0
0 Eq. 2
Alemdar
20
40
60
Park_87
80
100
120
Paulay
L (in.)
Mander
Figure 5 Comparisons of the plastic hinge length expressions results with Eq. 2
Lp/d -measured/Lp/d-estimated
Alemdar Eq. 1
4.5
Bae
Park 82
3.5
Park 87
Alemdar Eq. 1
2.5
Paulay 92
Bae
Mander 83
Berry 08
1.5
1
0.5
0
0
Alemdar Eq. 1
0.1
0.2
0.3
0.4
0.5
Bae
0.6
0.7
0.8
'
P/fc Ag (%)
Figure 6 Comparisons of the normalized plastic hinge length results with axial load level.
230
REFERENCES
231
232
18. Berry, M., Lehman D. E., and Lowes L. N. (2008), Lumped-Plasticity Models
for Performance Simulation of Bridge Columns, ACI Structural Journal, V. 105,
No. 3, May-June, pp. 270-279.
19. Hachem M. M., Mahin, S.A., and Moehle J.P. Performance of Circular
Reinforced Concrete Bridge Columns under Bidirectional Earthquake Loading.
Report No. PEER 2003/06, Pacific Earthquake Engineering Research Center,
University of California at Berkeley, Feb. 2003, 490 pp.
20. Dodd L.L., The Dynamic Behavior of Reinforced Concrete Bridge Piers
Subjected to New Zealand Seismicity, PhD. Thesis, the University of
Canterbury Christchurch New Zealand, Feb. 1992, 448 pages.
21. Dodd, L. L., and Cooke, N., Capacity of Circular Bridge Columns Subjected to
Base Excitation, ACI Structural Journal, Vol. 97, No.2, March-April 2000,
pp.297-308.
22. Kunnath, Sashi, K.; El-Bahy, Ashraf; Taylor, Andrew; and Stone, William,
Cumulative Seismic Damage of Reinforced Concrete Bridge Piers, Technical
Report NCEER-97-0006, National Center for Earthquake Engineering Research,
September 1997, 147 pp.
23. Cheok, G.S.; and Stone, William C., Behavior of 1/6-Scale Model Bridge
Columns Subjected to Cycle Inelastic Loading, NBSIR 86-3494, Center for
Building Technology, National Engineering Laboratory, National Institute of
Standards and Technology, Gaithersburg.
24. Stone, William C.; and Cheok, Geraldine S., Inelastic Behavior of Full-Scale
Bridge Columns Subjected to Cyclic Loading, NIST BSS 166, Building Science
233
234
CONCLUSIONS
This study focused on the plastic hinging behavior of reinforced concrete bridge
columns. A review of the literature revealed that most of the tests that have explored this
behavior have developed calibrated expressions to estimate regions of lumped plasticity,
which are then used to estimate maximum drift. There are a fewer number of tests that
have actually measured the spread of plasticity. In addition, the number of tested
elements that are representative of bridge columns (having lower axial load, lower
reinforcement ratios, and reasonable concrete compressive strengths) is limited. This
study sought to better define the plastic hinging behavior of reinforced concrete bridge
columns by taking detailed measurements of the column deformations in a large-scale 4span bridge test at the University of Nevada Reno, modeling the column using finite
elements, and developing separate expressions to calculate the maximum drift and the
spread of plasticity.
Three manuscripts were developed in the course of this research study. The
conclusions from each manuscript are repeated below.
According to the FDE index calculations, the lateral deformations of the bridge
column reduced from the photogrammetric data gave very good correlations with
235
The rotation of a vertical line on the grid surface provided a good representation
of the column rotation.
The deformed shape of the bridge column was constructed using the lateral
displacements of the points on the bottom and top grid surfaces and matched
well with the images at the time of maximum drift of Test 4D.
Using a 1 in. (25 mm) mesh along the hinging regions of the bridge column
provided a similar force-displacement curve with the Modified Compression
Field Theory.
According to the FDE index analysis, using the yield strength of 80 ksi (552
MPa) gave the best correlations between the FE model and the traditional
instrumentations.
The flexural stiffness of the cap beam can be modeled by defining a rotational
spring with 5*10^8 lb-in. (56.5*10^6 N-m) capacity.
The tension and compression recovery variables were utilized to consider the
reduction in stiffness of the concrete.
The best FE model of the bridge column under dynamic loading was defined
using a stress-strain relationship for the longitudinal steel and concrete with the
dynamic magnification factors and including the tension stiffening effect.
236
The bond-slip effect was defined in the FE model to consider the slip of the
longitudinal steel at the column interfaces.
The FE model of the bridge column followed a similar deformed shape obtained
from the photogrammetry data.
Two different plastic hinge length expressions were developed using the plastic
hinge length test data available in the literature. Equation 1 was produced based
on the plastic hinge length values that were calibrated to calculate the correct
maximum drift at the column. Equation 2 was developed using the measured
plastic hinge length values based on the spread of plastic curvature along the
height of the column.
The proposed equations gave the best fit at low axial loads.
The maximum drift capacity of the bridge column was estimated using the
proposed equations. The results show that the proposed equations give very good
estimates of the maximum drift capacity for the bridge column under dynamic
loading.
The plastic hinge length calculated using Eq. 1 gives shorter in value than that of
Eq. 2. More parametric analysis is required to evaluate the effect of the spread of
plasticity.
The study of plastic hinging regions for reinforced concrete bridge columns
shows that best estimates of the plastic hinge length depends on the goal of the analysis:
237
either for estimating the maximum drift capacity, or the spread of plasticity (>y). In
either case, it was found that plastic deformations can be measured with a simple
photogrammetry method during dynamic tests. The plastic behavior of the column was
also adequately represented using finite element modeling. Separate equations were
developed to represent the plastic hinge length for a lumped plasticity model (to estimate
the maximum drift) and for the spread of plasticity. The developed expressions provide
better estimates of plastic hinge lengths for lumped plasticity models at low axial loads
than any other expression that was considered. The proposed expression for a lumped
plasticity model was used effectively to estimate the maximum drift capacity. More
parametric analysis is needed, however, to understand the behavior of the spread of
plasticity.
238
ACI-ASCE Committee 428 (1968). Progress Report on Code Clauses for Limit
Design, ACI Journal, 65, No.9, pp. 713-715.
ACI Committee 318 (1989). Building Code Requirements for Reinforced Concrete and
Commentary (ACI 318-89/ACI 318R-89), American Concrete Institute, Detroit,
Michigan, 351 pp.
ACI Committee 318 (2005). Building Code Requirements for Reinforced Concrete and
Commentary (ACI 318-05/ACI 318R-05), American Concrete Institute, Detroit,
Michigan, 423 pp.
Ang Beng Ghee, Priestley M. J. N., and Park R. (1981). Ductility of Reinforced
Concrete Bridge Piers under Seismic Loading, Department of Civil Engineering
Research Report 81-3, University of Canterbury.
Bae, S. and Bayrak O. Plastic Hinge Length of Reinforced Concrete Columns, ACI
Structural Journal, V. 105, N. 3, May-June 2008, pp. 290-300.
239
Baker, A.L.L. (1956). Ultimate Load Theory Applied to the Design of Reinforced and
Prestressed Concrete Frames, Concrete Publications Ltd., London.
Baker, A.L.L. and Amarakone, A.M.N. (1964). Inelastic Hyperstatic Frame Analysis,
Proceedings International Symposium on the Flexural Mechanics of Reinforced
Concrete, ACI SP-12, Miami, pp. 85-142.
Bayrak, O. and Sheikh, S.A. (1997). High-Strength Concrete Columns under Simulated
Earthquake Loading, ACI Structural Journal, V. 94, No. 6, Nov.-Dec., pp. 708-722.
Berry, M., Parish, M., and Eberhard, M. (2004). PEER Structural Performance
Database Users Manual, PEER Research Report, University of California-Berkeley,
Berkeley, CA, 2004, http://nisee.berkeley.edu/spd.
Berry, M., Lehman D. E., and Lowes L. N. (2008). Lumped-Plasticity Models for
Performance Simulation of Bridge Columns, ACI Structural Journal, V. 105, No. 3,
May-June, pp. 270-279.
240
Davey B. E., and Park R. (1975). Reinforced Concrete Bridge Piers Under Seismic
Loading, Department of Civil Engineering Research Report 75-3, University of
Canterbury.
Dodd L.L. (1992). The Dynamic Behavior of Reinforced Concrete Bridge Piers
Subjected to New Zealand Seismicity, PhD. Thesis, the University of Canterbury
Christchurch New Zealand, 448 pages.
Dodd, L. L., and Cooke, N. (2000). Capacity of Circular Bridge Columns Subjected to
Base Excitation, ACI Structural Journal, Vol. 97, No.2, pp.297-308.
Ernst G. C. (1957). Plastic Hinging at the Intersection of Beams and Columns, Journal
of American Concrete Institute, Vol. 28, No.12, pp.1119-1144.
241
Gill W. D., Park R., and Priestley M. J. N. (1979). Ductility of Rectangular Reinforced
Concrete Columns with Axial Load, Department of Civil Engineering Research Report
79-1, University of Canterbury.
Hachem M. M., Mahin, S.A., and Moehle J.P. (2003). Performance of Circular
Reinforced Concrete Bridge Columns under Bidirectional Earthquake Loading, Report
No. PEER 2003/06, Pacific Earthquake Engineering Research Center, University of
California at Berkeley, 490 pp.
Mander J. B., Priestley M. J. N., and Park R. (1984). Seismic Design of Bridge Piers,
Department of Civil Engineering Research Report 84-2, University of Canterbury, 483
pp.
242
Munro I. R. M., Priestley M. J. N., and Park R. (1976). Seismic Behavior of Reinforced
Concrete Bridge Piers, Department of Civil Engineering Research Report 76-7,
University of Canterbury.
Ng Kit Heng, Priestley M. J. N., and Park R. (1978). Seismic Behavior of Circular
Reinforced Concrete Bridge Piers, Department of Civil Engineering Research Report
78-14, University of Canterbury.
243
Paulay T., and Priestley M. J. N. (1992). Seismic Design of Reinforced Concrete and
Masonry Structures, John Wiley & Sons, Inc.
Phan V., Saiidi M.S., Anderson J., and Ghasemi H. (2007). Near-Fault Ground Motion
Effects on Reinforced Concrete Bridge Columns, Journal of Structural Eng., V.133,
Issue 7, pp. 982-989.
Priestley, M. J. N., and Park R. (1987). Strength and Ductility of Concrete Bridge
Columns under Seismic Loading, ACI Structural Journal, V.84, Issue 1, pp. 61-76.
244
Restrepo J. I., Seible F., Stephan B., and Schoettler J. M. (2006). Seismic Testing of
Bridge Columns Incorporating High-Performance Materials, ACI Structural Journal,
V.103, Issue 4, pp. 496-504.
RRU Project 4805 (1983). Ductility of Reinforced Concrete Bridge Piers Under
Seismic Loading, Road Research Unit Newsletter, RRU NL 75, National Roads Board,
Wellington, New Zealand, pp.27-28.
Sheikh, S.A. and Khoury, S.S. (1993). Confined Concrete Columns with Stubs, ACI
Structural Journal, V. 90, No. 4, July-Aug., pp. 414-431.
Sheikh, S.A., Shah, D.V. and Khoury, S.S. (1994). Confinement of High-Strength
Concrete Columns, ACI Structural Journal, V. 91, No. 1, Jan.-Feb., pp. 100-111.
Soesianawati, M.T., Park, R. and Priestley, M.J.N. (1986). Limited Ductility Design of
Reinforced Concrete Columns, Research Report 86-10, Department of Civil
Engineering, University of Canterbury, Christchurch, New Zealand, Mar., 208 pp.
Standards Association of New Zealand (1982). New Zealand Standard Code of Practice
for the Design of Concrete Structures, NZS 3101, Wellington, New Zealand.
245
Watson, S. and Park, R. (1994). Simulated Seismic Load Tests on Reinforced Concrete
Columns, Journal of Structural Engineering, ASCE, V. 120, No. 6, June, pp. 18251849.
246
APPENDIX A
Appendix A contains the comparisons of rotations at each LVDT location for the
three different tests as described in Chapter 2.
Rotation (rad)
0.005
0
0
10
15
20
25
30
-0.005
-0.01
-0.015
Time (sec)
247
35
Rotation (rad)
0.005
0
0
10
15
20
25
30
35
-0.005
-0.01
-0.015
Time (sec)
Rotation (rad)
0.01
0
0
10
15
20
25
30
-0.01
-0.02
-0.03
Time (sec)
248
35
Rotation (rad)
0.005
0
0
10
15
20
25
30
35
-0.005
-0.01
-0.015
Time (sec)
Cross-Sec. Rot.
Vertical Rot.
0.015
0.01
Rotation (rad)
0.005
0
-0.005
10
15
20
25
30
-0.01
-0.015
-0.02
-0.025
-0.03
Time (sec)
249
35
Rotation (rad)
0.01
0
0
10
15
20
25
30
35
-0.01
-0.02
-0.03
-0.04
Time (sec)
0.04
0.03
Rotation (rad)
0.02
0.01
0
0
10
15
20
25
30
-0.01
-0.02
-0.03
Time (sec)
250
35
Cross-sec. Rot.
0.03
Rotation (rad)
0.02
0.01
0
0
10
15
20
25
30
35
-0.01
-0.02
-0.03
Time (sec)
0.03
Rotation (rad)
0.02
0.01
0
0
10
15
20
25
30
-0.01
-0.02
-0.03
Time (sec)
251
35
0.05
0.04
Rotation (rad)
0.03
0.02
0.01
0
0
10
15
20
25
30
35
-0.01
-0.02
-0.03
Time (sec)
Rotation (rad)
0.04
0.02
0
0
10
15
20
25
30
-0.02
-0.04
-0.06
Time (sec)
252
35
Rotation (rad)
0.04
0.02
0
0
10
15
20
25
30
-0.02
-0.04
-0.06
Time (sec)
253
35
APPENDIX B
Note: Comparisons are done with the combined movements of DT7, DS5 and DS1
instrumentation. Note that there is difference of amplitude between the results because of
a difference in height at the points of measurements.
0.4
Photogrammetry
0.3
DT7 & (DS5+DS1) combined
components
0.2
Displacement (in.)
0.1
0
0
10
15
20
25
30
35
-0.1
-0.2
-0.3
-0.4
-0.5
-0.6
Time (sec)
Fig. B-1: Lateral movement of Point 2 for Test 2 (1 in. = 254 mm)
254
40
0.4
Photogrammetry
DT7 & (DS5+DS1) combined
components
Displacement (in.)
0.2
0
0
10
15
20
25
30
35
40
-0.2
-0.4
-0.6
Time (sec)
Fig. B-2: Lateral movement of Point 3 for Test 2 (1 in. = 254 mm)
0.4
Photogrammetry
DT7 & (DS5+DS1) combined
components
Displacement (in.)
0.2
0
0
10
15
20
25
30
35
-0.2
-0.4
-0.6
Time (sec)
Fig. B-3: Lateral movement of Point 4 for Test 2 (1 in. = 254 mm)
255
40
0.4
Photogrammetry
DT7 & (DS5+DS1) combined
components
Displacement (in.)
0.2
0
0
10
15
20
25
30
35
40
-0.2
-0.4
-0.6
Time (sec)
Fig. B-4: Lateral movement of Point 7 for Test 2 (1 in. = 254 mm)
0.4
Photogrammetry
DT7 & (DS5+DS1) combined
components
Displacement (in.)
0.2
0
0
10
15
20
25
30
35
-0.2
-0.4
-0.6
Time (sec)
Fig. B-5: Lateral movement of Point 8 for Test 2 (1 in. = 254 mm)
256
40
0.4
Photogrammetry
DT7 & (DS5+DS1) combined
components
Displacement (in.)
0.2
0
0
10
15
20
25
30
35
40
-0.2
-0.4
-0.6
Time (sec)
Fig. B-6: Lateral movement of Point 9 for Test 2 (1 in. = 254 mm)
0.4
Photogrammetry
DT7 & (DS5+DS1) combined
components
Displacement (in.)
0.2
0
0
10
15
20
25
30
35
-0.2
-0.4
-0.6
Time (sec)
Fig. B-7: Lateral movement of Point 12 for Test 2 (1 in. = 254 mm)
257
40
0.4
Photogrammetry
DT7 & (DS5+DS1) combined
components
Displacement (in.)
0.2
0
0
10
15
20
25
30
35
40
-0.2
-0.4
-0.6
Time (sec)
Fig. B-8: Lateral movement of Point 13 for Test 2 (1 in. = 254 mm)
0.4
Photogrammetry
DT7 & (DS5+DS1) combined
components
Displacement (in.)
0.2
0
0
10
15
20
25
30
35
-0.2
-0.4
-0.6
Time (sec)
Fig. B-9: Lateral movement of Point 14 for Test 2 (1 in. = 254 mm)
258
40
0.4
Photogrammetry
DT7 & (DS5+DS1) combined
components
Displacement (in.)
0.2
0
0
10
15
20
25
30
35
40
-0.2
-0.4
-0.6
Time (sec)
Fig. B-10: Lateral movement of Point 17 for Test 2 (1 in. = 254 mm)
0.4
Photogrammetry
DT7 & (DS5+DS1) combined
components
Displacement (in.)
0.2
0
0
10
15
20
25
30
35
-0.2
-0.4
-0.6
Time (sec)
Fig. B-11: Lateral movement of Point 18 for Test 2 (1 in. = 254 mm)
259
40
0.4
Photogrammetry
DT7 & (DS5+DS1) combined
components
Displacement (in.)
0.2
0
0
10
15
20
25
30
35
40
-0.2
-0.4
-0.6
Time (sec)
Fig. B-12: Lateral movement of Point 19 for Test 2 (1 in. = 254 mm)
0.8
Photogrammetry
0.6
Displacement (in.)
0.4
0.2
0
0
10
15
20
25
30
35
-0.2
-0.4
-0.6
-0.8
Time (sec)
Fig. B-13: Lateral movement of Point 31 for Test 2 (1 in. = 254 mm)
260
40
0.8
Photogrammetry
DT7 & (DS5+DS1) combined
components
0.6
Displacement (in.)
0.4
0.2
0
0
10
15
20
25
30
35
40
-0.2
-0.4
-0.6
-0.8
Time (sec)
Fig. B-14: Lateral movement of Point 37 for Test 2 (1 in. = 254 mm)
0.8
Photogrammetry
DT7 & (DS5+DS1) combined
components
0.6
Displacement (in.)
0.4
0.2
0
0
10
15
20
25
30
35
-0.2
-0.4
-0.6
-0.8
Time (sec)
Fig. B-15: Lateral movement of Point 38 for Test 2 (1 in. = 254 mm)
261
40
0.8
Photogrammetry
0.6
Displacement (in.)
0.4
0.2
0
0
10
15
20
25
30
35
40
-0.2
-0.4
-0.6
-0.8
Time (sec)
Fig. B-16: Lateral movement of Point 39 for Test 2 (1 in. = 254 mm)
0.8
Photogrammetry
DT7 & (DS5+DS1) combined
components
0.6
Displacement (in.)
0.4
0.2
0
0
10
15
20
25
30
35
-0.2
-0.4
-0.6
-0.8
Time (sec)
Fig. B-17: Lateral movement of Point 44 for Test 2 (1 in. = 254 mm)
262
40
0.8
Photogrammetry
DT7 & (DS5+DS1) combined
components
0.6
Displacement (in.)
0.4
0.2
0
0
10
15
20
25
30
35
40
-0.2
-0.4
-0.6
-0.8
Time (sec)
Fig. B-18: Lateral movement of Point 45 for Test 2 (1 in. = 254 mm)
0.8
Photogrammetry
DT7 & (DS5+DS1) combined
components
0.6
Displacement (in.)
0.4
0.2
0
0
10
15
20
25
30
35
-0.2
-0.4
-0.6
-0.8
Time (sec)
Fig. B-19: Lateral movement of Point 46 for Test 2 (1 in. = 254 mm)
263
40
1
Photogrammetry
0.8
0.6
Displacement (in.)
0.4
0.2
0
0
10
15
20
25
30
35
40
-0.2
-0.4
-0.6
-0.8
-1
Time (sec)
Fig. B-20: Lateral movement of Point 51 for Test 2 (1 in. = 254 mm)
0.8
Photogrammetry
DT7 & (DS5+DS1) combined
components
0.6
Displacement (in.)
0.4
0.2
0
0
10
15
20
25
30
35
-0.2
-0.4
-0.6
-0.8
Time (sec)
Fig. B-21: Lateral movement of Point 52 for Test 2 (1 in. = 254 mm)
264
40
1
Photogrammetry
DT7 & (DS5+DS1) combined
components
0.8
0.6
Displacement (in.)
0.4
0.2
0
0
10
15
20
25
30
35
40
-0.2
-0.4
-0.6
-0.8
Time (sec)
Fig. B-22: Lateral movement of Point 53 for Test 2 (1 in. = 254 mm)
1
Photogrammetry
0.8
0.6
Displacement (in.)
0.4
0.2
0
0
10
15
20
25
30
35
-0.2
-0.4
-0.6
-0.8
Time (sec)
Fig. B-23: Lateral movement of Point 58 for Test 2 (1 in. = 254 mm)
265
40
0.8
Photogrammetry
0.6
0.4
Displacement (in.)
0.2
0
0
10
15
20
25
30
35
40
-0.2
-0.4
-0.6
-0.8
-1
Time (sec)
Fig. B-24: Lateral movement of Point 59 for Test 2 (1 in. = 254 mm)
1
Photogrammetry
0.8
0.6
Displacement (in.)
0.4
0.2
0
0
10
15
20
25
30
35
-0.2
-0.4
-0.6
-0.8
Time (sec)
Fig. B-25: Lateral movement of Point 60 for Test 2 (1 in. = 254 mm)
266
40
1.5
Photogrammetry
DT7 & (DS5+DS1) combined
components
Displacement (in.)
0.5
0
0
10
15
20
25
30
35
40
-0.5
-1
-1.5
-2
Time (sec)
Fig. B-26: Lateral movement of Point 2 for Test 4D (1 in. = 254 mm)
1.5
Photogrammetry
DT7 & (DS5+DS1) combined
components
1.0
Displacement (in.)
0.5
0.0
0
10
15
20
25
30
-0.5
-1.0
-1.5
-2.0
Time (sec)
Fig. B-27: Lateral movement of Point 3 for Test 4D (1 in. = 254 mm)
267
35
1.5
Photogrammetry
DT7 & (DS5+DS1) combined
components
1.0
Displacement (in.)
0.5
0.0
0
10
15
20
25
30
35
40
-0.5
-1.0
-1.5
-2.0
Time (sec)
Fig. B-28: Lateral movement of Point 4 for Test 4D (1 in. = 254 mm)
1.5
Photogrammetry
DT7 & (DS5+DS1) combined
components
1.0
Displacement (in.)
0.5
0.0
0
10
15
20
25
30
35
-0.5
-1.0
-1.5
-2.0
Time (sec)
Fig. B-29: Lateral movement of Point 7 for Test 4D (1 in. = 254 mm)
268
40
1.5
Photogrammetry
DT7 & (DS5+DS1) combined
components
1.0
Displacement (in.)
0.5
0.0
0
10
15
20
25
30
35
40
-0.5
-1.0
-1.5
-2.0
Time (sec)
Fig. B-30: Lateral movement of Point 8 for Test 4D (1 in. = 254 mm)
1.5
Photogrammetry
DT7 & (DS5+DS1) combined
components
1.0
Displacement (in.)
0.5
0.0
0
10
15
20
25
30
35
-0.5
-1.0
-1.5
-2.0
Time (sec)
Fig. B-31: Lateral movement of Point 9 for Test 4D (1 in. = 254 mm)
269
40
1.5
Photogrammetry
DT7 & (DS5+DS1) combined
components
Displacement (in.)
0.5
0
0
10
15
20
25
30
35
-0.5
-1
-1.5
-2
Time (sec)
Fig. B-32: Lateral movement of Point 12 for Test 4D (1 in. = 254 mm)
1.5
Photogrammetry
DT7 & (DS5+DS1) combined
components
1.0
Displacement (in.)
0.5
0.0
0
10
15
20
25
30
-0.5
-1.0
-1.5
-2.0
Time (sec)
Fig. B-33: Lateral movement of Point 13 for Test 4D (1 in. = 254 mm)
270
35
1.5
Photogrammetry
DT7 & (DS5+DS1) combined
components
1.0
Displacement (in.)
0.5
0.0
0
10
15
20
25
30
35
40
-0.5
-1.0
-1.5
-2.0
Time (sec)
Fig. B-34: Lateral movement of Point 14 for Test 4D (1 in. = 254 mm)
1.5
Photogrammetry
DT7 & (DS5+DS1) combined
components
Displacement (in.)
0.5
0
0
10
15
20
25
30
-0.5
-1
-1.5
-2
Time (sec)
Fig. B-35: Lateral movement of Point 17 for Test 4D (1 in. = 254 mm)
271
35
1.5
Photogrammetry
DT7 & (DS5+DS1) combined
components
1.0
Displacement (in.)
0.5
0.0
0
10
15
20
25
30
35
40
-0.5
-1.0
-1.5
-2.0
Time (sec)
Fig. B-36: Lateral movement of Point 18 for Test 4D (1 in. = 254 mm)
1.5
Photogrammetry
DT7 & (DS5+DS1) combined
components
1.0
Displacement (in.)
0.5
0.0
0
10
15
20
25
30
35
-0.5
-1.0
-1.5
-2.0
Time (sec)
Fig. B-37: Lateral movement of Point 19 for Test 4D (1 in. = 254 mm)
272
40
2.5
Photogrammetry
2
Displacement (in.)
1.5
0.5
0
0
10
15
20
25
30
35
-0.5
-1
-1.5
Time (sec)
Fig. B-38: Lateral movement of Point 37 for Test 4D (1 in. = 254 mm)
2.5
Photogrammetry
2
DT7 & (DS5+DS1) combined
components
Displacement (in.)
1.5
0.5
0
0
10
15
20
25
30
-0.5
-1
-1.5
Time (sec)
Fig. B-39: Lateral movement of Point 38 for Test 4D (1 in. = 254 mm)
273
35
2.5
Photogrammetry
DT7 & (DS5+DS1) combined
components
Displacement (in.)
1.5
0.5
0
0
10
15
20
25
30
-0.5
-1
-1.5
Time (sec)
Fig. B-40: Lateral movement of Point 39 for Test 4D (1 in. = 254 mm)
2.5
Photogrammetry
2
Displacement (in.)
1.5
0.5
0
0
10
20
30
-0.5
-1
-1.5
Time (sec)
Fig. B-41: Lateral movement of Point 44 for Test 4D (1 in. = 254 mm)
274
35
2.5
Photogrammetry
DT7 & (DS5+DS1) combined
components
Displacement (in.)
1.5
0.5
0
0
10
20
30
-0.5
-1
-1.5
Time (sec)
Fig. B-42: Lateral movement of Point 45 for Test 4D (1 in. = 254 mm)
3
Photogrammetry
2.5
Displacement (in.)
1.5
0.5
0
0
10
20
30
-0.5
-1
-1.5
Time (sec)
Fig. B-43: Lateral movement of Point 46 for Test 4D (1 in. = 254 mm)
275
2.5
Photogrammetry
2
Displacement (in.)
1.5
0.5
0
0
10
20
30
-0.5
-1
-1.5
Time (sec)
Fig. B-44: Lateral movement of Point 51 for Test 4D (1 in. = 254 mm)
2.5
Photogrammetry
DT7 & (DS5+DS1) combined
components
Displacement (in.)
1.5
0.5
0
0
10
15
20
25
30
-0.5
-1
-1.5
Time (sec)
Fig. B-45: Lateral movement of Point 52 for Test 4D (1 in. = 254 mm)
276
35
2.5
Photogrammetry
2
Displacement (in.)
1.5
0.5
0
0
10
15
20
25
30
35
-0.5
-1
-1.5
Time (sec)
Fig. B-46: Lateral movement of Point 53 for Test 4D (1 in. = 254 mm)
2.5
Photogrammetry
DT7 & (DS5+DS1) combined
components
Displacement (in.)
1.5
0.5
0
0
10
15
20
25
30
-0.5
-1
-1.5
Time (sec)
Fig. B-47: Lateral movement of Point 58 for Test 4D (1 in. = 254 mm)
277
35
2.5
Photogrammetry
DT7 & (DS5+DS1) combined
components
Displacement (in.)
1.5
0.5
0
0
10
15
20
25
30
35
40
-0.5
-1
-1.5
Time (sec)
Fig. B-48: Lateral movement of Point 59 for Test 4D (1 in. = 254 mm)
2.5
Photogrammetry
2
Displacement (in.)
1.5
0.5
0
0
10
15
20
25
30
-0.5
-1
-1.5
Time (sec)
Fig. B-49: Lateral movement of Point 60 for Test 4D (1 in. = 254 mm)
278
35
2.0
Photogrammetry
1.5
1.0
Displacement (in.)
0.5
0.0
0
10
15
20
25
30
35
40
-0.5
-1.0
-1.5
-2.0
-2.5
-3.0
-3.5
Time (sec)
Fig. B-50: Lateral movement of Point 2 for Test 6 (1 in. = 254 mm)
2.0
Photogrammetry
1.5
1.0
Displacement (in.)
0.5
0.0
0
10
15
20
25
30
35
-0.5
-1.0
-1.5
-2.0
-2.5
-3.0
-3.5
Time (sec)
Fig. B-51: Lateral movement of Point 3 for Test 6 (1 in. = 254 mm)
279
40
2.0
Photogrammetry
1.5
1.0
0.5
Displacement (in.)
0.0
0
10
15
20
25
30
35
40
-0.5
-1.0
-1.5
-2.0
-2.5
-3.0
-3.5
Time (sec)
Fig. B-52: Lateral movement of Point 4 for Test 6 (1 in. = 254 mm)
2.0
Photogrammetry
1.5
1.0
Displacement (in.)
0.5
0.0
0
10
15
20
25
30
35
-0.5
-1.0
-1.5
-2.0
-2.5
-3.0
-3.5
Time (sec)
Fig. B-53: Lateral movement of Point 7 for Test 6 (1 in. = 254 mm)
280
40
2.0
Photogrammetry
1.5
1.0
0.5
Displacement (in.)
0.0
0
10
15
20
25
30
35
40
-0.5
-1.0
-1.5
-2.0
-2.5
-3.0
-3.5
Time (sec)
Fig. B-54: Lateral movement of Point 8 for Test 6 (1 in. = 254 mm)
2.0
Photogrammetry
1.5
1.0
0.5
Displacement (in.)
0.0
0
10
15
20
25
30
35
-0.5
-1.0
-1.5
-2.0
-2.5
-3.0
-3.5
Time (sec)
Fig. B-55: Lateral movement of Point 9 for Test 6 (1 in. = 254 mm)
281
40
2.0
Photogrammetry
1.5
1.0
0.5
Displacement (in.)
0.0
0
10
15
20
25
30
35
40
-0.5
-1.0
-1.5
-2.0
-2.5
-3.0
-3.5
Time (sec)
Fig. B-56: Lateral movement of Point 12 for Test 6 (1 in. = 254 mm)
2.0
Photogrammetry
1.5
1.0
0.5
Displacement (in.)
0.0
0
10
15
20
25
30
35
-0.5
-1.0
-1.5
-2.0
-2.5
-3.0
-3.5
Time (sec)
Fig. B-57: Lateral movement of Point 13 for Test 6 (1 in. = 254 mm)
282
40
2.0
Photogrammetry
1.5
1.0
0.5
Displacement (in.)
0.0
0
10
15
20
25
30
35
40
-0.5
-1.0
-1.5
-2.0
-2.5
-3.0
-3.5
Time (sec)
Fig. B-58: Lateral movement of Point 14 for Test 6 (1 in. = 254 mm)
2.0
Photogrammetry
1.5
1.0
0.5
Displacement (in.)
0.0
0
10
15
20
25
30
35
-0.5
-1.0
-1.5
-2.0
-2.5
-3.0
-3.5
Time (sec)
Fig. B-59: Lateral movement of Point 17 for Test 6 (1 in. = 254 mm)
283
40
2.0
Photogrammetry
1.5
1.0
0.5
Displacement (in.)
0.0
0
10
15
20
25
30
35
40
-0.5
-1.0
-1.5
-2.0
-2.5
-3.0
-3.5
Time (sec)
Fig. B-60: Lateral movement of Point 18 for Test 6 (1 in. = 254 mm)
2.0
Photogrammetry
1.5
DT7 & (DS5+DS1) combined
components
1.0
Displacement (in.)
0.5
0.0
0
10
15
20
25
30
35
-0.5
-1.0
-1.5
-2.0
-2.5
-3.0
-3.5
Time (sec)
Fig. B-61: Lateral movement of Point 19 for Test 6 (1 in. = 254 mm)
284
40
5.00
Photogrammetry
4.00
Displacement (in.)
3.00
2.00
1.00
0.00
0
10
15
20
25
30
35
40
-1.00
-2.00
-3.00
Time (sec)
Fig. B-62: Lateral movement of Point 37 for Test 6 (1 in. = 254 mm)
5.00
Photogrammetry
DT7 & (DS5+DS1) combined
components
4.00
Displacement (in.)
3.00
2.00
1.00
0.00
0
10
15
20
25
30
35
-1.00
-2.00
-3.00
Time (sec)
Fig. B-63: Lateral movement of Point 38 for Test 6 (1 in. = 254 mm)
285
40
5.00
Photogrammetry
4.00
Displacement (in.)
3.00
2.00
1.00
0.00
0
10
15
20
25
30
35
40
-1.00
-2.00
-3.00
Time (sec)
Fig. B-64: Lateral movement of Point 39 for Test 6 (1 in. = 254 mm)
5.00
Photogrammetry
DT7 & (DS5+DS1) combined
components
4.00
Displacement (in.)
3.00
2.00
1.00
0.00
0
10
15
20
25
30
35
-1.00
-2.00
-3.00
Time (sec)
Fig. B-65: Lateral movement of Point 44 for Test 6 (1 in. = 254 mm)
286
40
5.00
Photogrammetry
4.00
Displacement (in.)
3.00
2.00
1.00
0.00
0
10
15
20
25
30
35
40
-1.00
-2.00
-3.00
Time (sec)
Fig. B-66: Lateral movement of Point 45 for Test 6 (1 in. = 254 mm)
5.00
Photogrammetry
DT7 & (DS5+DS1) combined
components
4.00
Displacement (in.)
3.00
2.00
1.00
0.00
0
10
15
20
25
30
35
-1.00
-2.00
-3.00
Time (sec)
Fig. B-67: Lateral movement of Point 46 for Test 6 (1 in. = 254 mm)
287
40
5.00
Photogrammetry
DT7 & (DS5+DS1) combined
components
4.00
Displacement (in.)
3.00
2.00
1.00
0.00
0
10
15
20
25
30
35
40
-1.00
-2.00
-3.00
Time (sec)
Fig. B-68: Lateral movement of Point 51 for Test 6 (1 in. = 254 mm)
5.00
Photogrammetry
DT7 & (DS5+DS1) combined
components
4.00
Displacement (in.)
3.00
2.00
1.00
0.00
0
10
15
20
25
30
35
-1.00
-2.00
-3.00
Time (sec)
Fig. B-69: Lateral movement of Point 52 for Test 6 (1 in. = 254 mm)
288
40
5.00
Photogrammetry
DT7 & (DS5+DS1) combined
components
4.00
3.00
Displacement (in.)
2.00
1.00
0.00
0
10
15
20
25
30
35
40
-1.00
-2.00
-3.00
Time (sec)
Fig. B-70: Lateral movement of Point 53 for Test 6 (1 in. = 254 mm)
5.00
Photogrammetry
DT7 & (DS5+DS1) combined
components
4.00
Displacement (in.)
3.00
2.00
1.00
0.00
0
10
15
20
25
30
35
-1.00
-2.00
-3.00
Time (sec)
Fig. B-71: Lateral movement of Point 58 for Test 6 (1 in. = 254 mm)
289
40
5.0
Photogrammetry
4.0
Displacement (in.)
3.0
2.0
1.0
0.0
0
10
15
20
25
30
35
40
-1.0
-2.0
-3.0
Time (sec)
Fig. B-72: Lateral movement of Point 59 for Test 6 (1 in. = 254 mm)
5.00
Photogrammetry
DT7 & (DS5+DS1) combined
components
4.00
Displacement (in.)
3.00
2.00
1.00
0.00
0
10
15
20
25
30
35
-1.00
-2.00
-3.00
Time (sec)
Fig. B-73: Lateral movement of Point 60 for Test 6 (1 in. = 254 mm)
290
40